Frequently Asked Questions

Pirai AI accepts Word documents (.docx) and PDF files. Whether you have a native Word document, a digital PDF, or any other document format, simply upload it and Pirai AI will automatically extract and structure all your survey content. The system preserves all text in its original language and formatting, regardless of the file type.

Yes! Pirai AI can process scanned documents and even photos of paper surveys. If you have a paper questionnaire or printed survey, simply take a photo with your phone, convert it to PDF, and upload it to Pirai AI. Our advanced extraction technology reads questions, response options, tables, and all survey elements from images and scanned documents with the same accuracy as digital files. This makes it easy to digitize legacy surveys, research instruments, or any paper-based questionnaires.

Absolutely! Pirai AI is designed to process surveys of any size and complexity without compromising accuracy. Whether you have a 10-question survey or a comprehensive 200-question research instrument, Pirai AI delivers complete, accurate conversions in the same amount of time — typically under 10 minutes. The system automatically adapts to your document's size and structure, ensuring every question, response option, and matrix element is captured perfectly.

No! Pirai AI adapts to your existing document style. Whether you use "Q1. How satisfied..." or "1) Rate your satisfaction..." or any other formatting, the AI understands your structure. You can keep using your current templates without any changes.

Pirai AI automatically checks for spelling errors across all questions, answer options, and text content during conversion. The system generates a comprehensive typo report that highlights potential misspellings with their locations. You can download the converted Excel file, review and fix any typos or issues identified, and then upload the edited version in Step 2 for platform conversion. The Excel template structure is designed to be edit-friendly — you can make corrections directly in Excel without breaking the formatting, and Pirai AI will process your edited file seamlessly when converting to any survey platform.

Yes! Pirai AI supports surveys in 50+ languages with UTF-8 compliant processing. All text is preserved in its original script and language without translation or transliteration. This includes French, German, Portuguese, Spanish, Italian, Arabic, Chinese, Japanese, and many more.

The Excel template uses 7 standardized columns with exact headers that must not be modified: Block Name | Question ID | Question Text | Question Type | Answer Option Text | Recode Value | Matrix Type. This structure captures all survey elements in a platform-agnostic format that can be converted to 20+ target platforms. Data entry starts from Row 2, as Row 1 contains the mandatory column headers.

Block Name (Column A): Groups questions into survey sections. Required for each new question with consistent naming like 'Block 1' or 'Demographics'. Question ID (Column B): Unique alphanumeric identifier for each question (e.g., 'Q1', 'DEMO_AGE'). Question Text (Column C): The actual question respondents see, required for new questions, empty for continuation rows. Question Type (Column D): Defines question behavior using exact type codes, empty for option rows. Answer Option Text (Column E): Choice text, statements, or labels varying by question type. Recode Value (Column F): Numeric codes for data analysis, required for most options. Matrix Type (Column G): Specifies 'Statement' or 'Scale Point' only for Matrix questions.

The column headers must match exactly as specified: Block Name | Question ID | Question Text | Question Type | Answer Option Text | Recode Value | Matrix Type. Any modification to these headers will break the conversion process and cause the system to fail. The structure integrity is critical for successful conversion — this is a mandatory requirement, not a guideline.

Continuation rows are used for answer options or additional elements of a question. For standard multiple choice questions: the first row contains Block Name, Question ID, Question Text, Question Type, first Answer Option, and Recode Value. Subsequent rows for the same question only fill in Block Name, Question ID, Answer Option Text, and Recode Value — leaving Question Text and Question Type cells empty. This pattern applies to most multi-option question types except Matrix questions which have special rules.

Block Name: Must be filled for every row including continuation rows. Use consistent naming to group related questions logically (e.g., 'Block 1', 'Demographics', 'Satisfaction Questions'). Question ID: Must be unique across the entire survey with no duplicates. Use alphanumeric characters only (e.g., 'Q1', 'Q2', 'DEMO_AGE'). The same Question ID must be repeated for all rows belonging to that question, including all answer options. Stick to clear, consistent naming conventions — sequential numbering works best.

Recode Values are numeric or string codes used for data analysis. For most question types, each answer option requires a unique recode value (typically sequential: 1, 2, 3…). For platforms that support custom recode values (like Qualtrics, REDCap, Forsta), you can use descriptive codes. For platforms without recode support, the system uses simplified sequential numbering. Text Entry questions typically leave Recode Value empty. Matrix questions assign recode values to both statements and scale points independently.

Before converting your Excel template, verify these mandatory requirements: (1) All 7 columns present with exact headers — no modifications allowed. (2) Every question has required fields filled — no gaps in critical data. (3) All answer options follow the correct pattern for their question type. (4) Matrix questions have statements listed before scale points with correct Matrix Type values. (5) No duplicate Question IDs exist within your survey. (6) Questions are grouped logically into blocks. (7) No problematic special characters in critical fields. (8) Test import a small sample successfully before building the full survey.

Common mistakes include: (1) Modified column headers — headers must match exactly as specified. (2) Missing empty cells in continuation rows — Question Text and Question Type must be empty for option rows. (3) Wrong Matrix order — statements must be listed before scale points. (4) Duplicate Question IDs — each question needs a unique identifier. (5) Inconsistent Block Names — use the exact same block name across related questions. (6) Special characters in Question IDs — stick to alphanumeric characters only. (7) Missing required fields — ensure all mandatory columns are filled for the first row of each question.

Use the Block Name column to organize your survey logically: (1) Group related questions together (e.g., all satisfaction questions in 'Satisfaction Block'). (2) Use consistent naming conventions across the survey. (3) Common block patterns include: Introduction, Main Questions (Block 1, Block 2, etc.), Demographics, Thank You. (4) Every row, including continuation rows, must have a Block Name filled in. (5) Block names appear in the converted survey platform as section headers or organizational units.

Yes! The Excel template is designed to be edit-friendly. After Step 1 (Word/PDF to Excel conversion), you can: (1) Review and fix any typos or spelling errors identified in the typo report. (2) Adjust question wording or answer options. (3) Add or remove questions following the correct structural patterns. (4) Modify recode values. (5) Reorganize blocks. The key is to maintain the exact 7-column structure and follow the correct patterns for each question type. You can then upload your edited Excel file in Step 2 for platform conversion. This makes Pirai AI ideal for team collaboration and review workflows.

For Question IDs, use only alphanumeric characters — avoid special characters like @, #, $, %, &, *, spaces, or punctuation. For Question Text and Answer Options, you can use normal punctuation and special characters, but be cautious with characters that might have special meaning in survey platforms. For Block Names, stick to simple alphanumeric names with spaces allowed. For Recode Values, use integers or simple strings without special characters. UTF-8 encoding is supported for international characters and multilingual content.

Follow this testing workflow: (1) Start small — create a sample with 2–3 questions covering different question types you plan to use. (2) Verify structure — ensure all 7 columns have correct headers and data follows exact patterns. (3) Test conversion — convert your sample in Step 2 to your target platform. (4) Import and review — import the generated file into your survey platform and verify questions display correctly. (5) Iterate — if issues arise, refer to examples and troubleshooting guides, fix the structure, and re-test. (6) Scale up — once your sample converts successfully, build your full survey following the same patterns.

Question Type Code: MC:SingleAnswer. Use Case: When respondents select only one option from a list. First row contains Block Name, Question ID, Question Text, Question Type='MC:SingleAnswer', first Answer Option Text, and Recode Value (typically 1). Subsequent rows contain Block Name, Question ID (repeated), empty Question Text, empty Question Type, next Answer Option Text, and sequential Recode Value. Rules: Minimum 2 options required; each option needs a unique recode value; continuation rows leave Question Text and Question Type empty.

Question Type Code: MC:MultipleAnswer. Use Case: When respondents can select multiple options (select all that apply). Structure pattern is identical to MC:SingleAnswer — first row has all question details with Question Type='MC:MultipleAnswer'. Continuation rows contain only Block Name, Question ID, Answer Option Text, and Recode Value. Consider using descriptive recode values for easier analysis (e.g., 'Option_A', 'Option_B').

Question Type Code: Matrix. Use Case: Grid-style questions with statements rated on a scale. First row contains Block Name, Question ID, Question Text, Question Type='Matrix', first Statement in Answer Option Text, Recode Value for statement (1), and Matrix Type='Statement'. Continue with ALL statements (Matrix Type='Statement'). Then list ALL scale points with Matrix Type='Scale Point'. CRITICAL: ALL statements must be listed FIRST, then ALL scale points — never intermix them. Question Text only appears in the first row; subsequent rows leave it empty.

Pirai AI supports three text entry types: TE:SingleLine for short responses — single row format with empty Answer Option Text, Recode Value, and Matrix Type. TE:Essay for long-form responses — identical single row structure. TE:Form for multiple input fields — first row contains question details; continuation rows contain Block Name, Question ID, and field labels in Answer Option Text (e.g., 'Name', 'Email', 'Phone'). Recode Value and Matrix Type remain empty for all text entry types.

Question Type Code: RankOrder. Use Case: When respondents need to rank items in order of preference. Standard multi-row format like multiple choice — first row contains Block Name, Question ID, Question Text, Question Type='RankOrder', first item, and Recode Value (1). Continuation rows contain Block Name, Question ID, item text, sequential recode values. Consider the number of items — too many (>7) can be overwhelming. Not all platforms support Rank Order; Pirai AI may convert to multiple choice for unsupported platforms with clear warnings.

Question Type Code: ConstantSum. Use Case: When responses must total a specific number, typically used for allocation questions (e.g., 'Allocate 100 points across these categories'). First row contains Question Text specifying the target sum, Question Type='ConstantSum', first category in Answer Option Text. Continuation rows list additional categories. Clearly specify the target sum in the question text, and use 3–7 categories for optimal user experience. The survey platform enforces the sum constraint.

Question Type Code: DB (Descriptive Block). Use Case: Display information without collecting responses — introductions, instructions, thank you messages, transitions, consent text. Structure Pattern: Single row only — Block Name, Question ID (e.g., 'INTRO', 'END'), Question Text (the message to display), Question Type='DB', and empty cells for Answer Option Text, Recode Value, and Matrix Type. No continuation rows needed. Question Text can contain HTML formatting on some platforms.

Question Type Code: Page Break. Use Case: Force a new page in the survey flow to control pacing, prevent survey fatigue, create natural breaks between topics. Structure Pattern: Single row with Block Name filled and all other columns empty except Question Type='Page Break'. Use strategically — recommended after every 5–7 questions or at natural topic transitions. Place demographics questions on a separate page at the end.

Common Matrix question mistakes: (1) Wrong Matrix Type order — the most critical error is listing scale points before statements. You MUST list ALL statements first, then ALL scale points. (2) Missing Matrix Type values — every row must specify either 'Statement' or 'Scale Point' in column G. (3) Intermixing statements and scales — never alternate between statements and scale points. (4) Wrong recode values — statements and scale points have independent recode sequences. (5) Forgetting continuation pattern — Question Text only appears in the first row. The pattern is: Question Text row → ALL Statements → ALL Scale Points.

Pirai AI supports 10 question types: MC:SingleAnswer, MC:MultipleAnswer, Matrix, TE:SingleLine, TE:Essay, TE:Form, RankOrder, ConstantSum, DB (Descriptive Block), and Page Break. All platforms support multiple choice and text entry. Matrix questions are natively supported by Qualtrics, QuestionPro, REDCap, Forsta, and Dimensions — for platforms without matrix support, Pirai AI automatically converts matrix questions into individual single-choice questions. Rank Order and Constant Sum have limited support and may be converted to alternative types on unsupported platforms with clear warnings.

Pirai AI supports 20+ platforms across market research, education, healthcare, and enterprise sectors, including Qualtrics (Advanced & QSF), REDCap, SurveyMonkey, QuestionPro, Forsta, Dimensions MDD, Zoho Surveys, SurveySparrow, Microsoft Forms, Google Forms, Moodle, Canvas, Kahoot, LearnWorlds, BrainCert, and more.

Yes, for platforms that support matrix questions natively (like Qualtrics, QuestionPro, REDCap, Forsta). For platforms without matrix support, Pirai AI automatically converts matrix questions into individual single-choice questions while preserving all statements and scale points.

Absolutely! After converting your Word document to the Excel template, you can convert that same Excel file to any of the 20+ supported platforms. This lets you deploy the same survey across multiple tools without rebuilding content from scratch.

Pirai AI supports: Multiple Choice (Single/Multiple Answer), Text Entry (Essay, Single Line, Form), Matrix questions, Rank Order, Constant Sum, Descriptive Blocks, Page Breaks, and platform-specific question types. The system intelligently converts unsupported question types to the closest equivalent for each platform.

Pirai AI provides intelligent conversion with warnings. For example, Rank Order questions may be converted to Multiple Choice for platforms without ranking support. Constant Sum questions might convert to text entry. You always receive clear notifications about any conversions made.

The typo check scans all question text, answer options, and matrix statements for potential spelling errors. It uses intelligent filtering to exclude technical terms, proper nouns, and domain-specific terminology. The report shows all potential typos with their locations and provides counts by question type.

Essential Do's: (1) Follow examples exactly — use the provided patterns as templates without deviation. (2) Use consistent Block Names — group related questions logically. (3) Sequential Question IDs — use clear, consistent naming (Q1, Q2, Q3, etc.). (4) Write clear Question Text — be concise and unambiguous to avoid respondent confusion. (5) Use meaningful Recode Values — implement logical coding systems for easier data analysis. (6) Test early and often — import small sections to verify structure before building the full survey. (7) Keep documentation — save incremental versions as you build complex surveys.

Critical Don'ts: (1) Don't modify column headers — this will break the conversion completely. (2) Don't skip required fields — empty required cells cause errors and failed conversions. (3) Don't mix question types — each question should have one consistent type throughout all its rows. (4) Don't use special characters in Question IDs — stick to alphanumeric characters only. (5) Don't forget Matrix order — statements MUST come before scale points. (6) Don't leave gaps — fill continuation rows properly. (7) Don't use duplicate Question IDs — each question needs a unique identifier across the entire survey.

Before conversion, complete this Quality Assurance Checklist: (1) Structure integrity — all 7 columns present with exact headers. (2) Question completeness — every question has all required fields filled in the first row. (3) Option consistency — all answer options follow the correct pattern for their question type. (4) Matrix format — statements listed before scale points with correct Matrix Type values. (5) Unique IDs — no duplicate Question IDs anywhere in the survey. (6) Block logic — questions grouped logically into meaningful blocks. (7) Special characters — no problematic characters in critical fields. (8) Test import — small sample successfully converts before building the full survey.

Efficient Workflow Strategies: (1) Plan your survey structure — sketch out blocks and question flow before entering any data in Excel. (2) Use copy-paste — duplicate similar questions and modify details rather than creating from scratch. (3) Batch similar questions — group all multiple choice questions together, all matrix questions together. (4) Version control — save incremental versions (e.g., Survey_v1, Survey_v2) as you build complex surveys. (5) Start with a template — create a master template with your standard question types and copy it for new surveys. (6) Review in stages — verify structure section by section rather than building the entire survey before checking.

Matrix Scale Design Best Practices: (1) Keep scales consistent — don't mix 5-point and 7-point scales in the same survey section. (2) Use 5–7 point scales for optimal discrimination. (3) Consider neutral midpoints — include 'Neither Agree nor Disagree' for topics where respondents may have no opinion. (4) Include Not Applicable — add 'N/A' options when appropriate. (5) Label endpoints clearly — always label both ends of the scale explicitly. (6) Use balanced scales — equal number of positive and negative options. Common scale types: Likert scales (agreement), frequency scales (Never to Always), satisfaction scales (Very Dissatisfied to Very Satisfied).

Question Ordering Strategy: (1) Start with engaging, easy questions — begin with interesting, non-threatening questions to build momentum. (2) Group related topics together — keep all questions about one subject in the same block. (3) Place demographic questions at the end — age, income, education can cause early dropouts if placed first. (4) Use page breaks to prevent fatigue — insert page breaks every 5–7 questions or between major topic sections. (5) Build from general to specific — start with broad questions, then drill down into details. (6) End with open-ended questions — place optional comment boxes at the end when respondents are most committed.

Review and Editing Process: (1) Download the Excel output from Step 1 and review the typo check report. (2) Check structural accuracy — verify all questions, options, and formatting are captured correctly. (3) Fix identified typos — correct spelling errors flagged in the report. (4) Adjust wording — refine question text and answer options while maintaining structure. (5) Share with stakeholders — the Excel format is perfect for team collaboration and approval workflows. (6) Maintain structure — when editing, never modify column headers. (7) Re-upload for conversion — once finalized, upload the edited Excel file in Step 2 to generate your platform file.

Success Indicators: (1) Structural precision — template exactly matches documented patterns. (2) Clean conversion — Step 2 conversion completes without errors or warnings. (3) Correct import — survey imports successfully into target platform. (4) Accurate display — all questions display correctly in the survey platform. (5) Valid logic — question types behave as expected. (6) Preserved data — all text, options, and recode values intact after conversion. (7) No duplicate IDs — each question has unique identification. (8) Logical flow — questions appear in intended order with appropriate page breaks. If these indicators are met, your Excel template is production-ready.

Problem: Conversion fails with 'Invalid structure' error. Solution: Check that column headers match exactly: Block Name | Question ID | Question Text | Question Type | Answer Option Text | Recode Value | Matrix Type. The headers are case-sensitive and must have this exact spelling and order. Do not add, remove, or modify any column headers. Verify that data entry starts from Row 2 (Row 1 contains headers). Check for any merged cells or hidden columns that might disrupt the structure. If headers are correct, verify that every row has a Block Name filled in.

Problem: Matrix question shows wrong format or missing elements. Solution: (1) Verify that all statements have Matrix Type = 'Statement' and all scale points have Matrix Type = 'Scale Point'. (2) Check that statements are listed FIRST, before any scale points — never intermix them. (3) Ensure the first row contains the question text in the Question Text column. (4) Confirm all rows except the first have Question Type = 'Matrix'. (5) Verify that both statements and scale points have sequential recode values. (6) Check for spelling errors in the Matrix Type column — it must be exactly 'Statement' or 'Scale Point' (case-sensitive). Critical: The order must be: Question Text row → ALL Statements → ALL Scale Points.

Problem: Multiple choice question shows only the first option or displays incorrectly. Solution: (1) Ensure continuation rows have empty cells in Question Text and Question Type columns — these should only be filled in the first row. (2) Verify that Block Name and Question ID are repeated in every row for that question. (3) Check that Answer Option Text is filled for each option. (4) Confirm each option has a unique Recode Value. (5) Verify there are no extra spaces or special characters in empty cells. (6) Check that Question Type in the first row exactly matches supported types: 'MC:SingleAnswer' or 'MC:MultipleAnswer' (case-sensitive). The pattern is: Row 1 = full question details, Row 2+ = Block Name + Question ID + Answer Option + Recode Value only.

Problem: The same question displays multiple times in the converted survey. Solution: (1) Check for duplicate Question IDs — each question must have a unique identifier. (2) Verify continuation rows — ensure Question Text and Question Type are empty in option rows. If these are filled in continuation rows, the system may interpret them as new questions. (3) Check for missing Block Name or Question ID in continuation rows. (4) Look for extra blank rows between question options that might break the question grouping. (5) Verify there are no hidden rows or filtered data that might cause issues.

Problem: Text Entry questions don't appear or display incorrectly. Solution: (1) Verify that Answer Option Text, Recode Value, and Matrix Type are empty for TE:SingleLine and TE:Essay types. (2) For TE:Form questions, check that continuation rows contain only Block Name, Question ID, and Answer Option Text (field labels). (3) Confirm the Question Type is spelled exactly as: TE:SingleLine, TE:Essay, or TE:Form (case-sensitive, with colon). (4) Ensure the question follows single-row format for SingleLine and Essay types — no continuation rows needed. (5) For Form type, verify each field label appears in separate continuation rows.

For assistance, questions, or support inquiries, please contact us at admin@piraiai.com. Our team is ready to help with any questions about conversions, platform compatibility, technical issues, Excel template structure, question type patterns, enterprise pricing, or agency support.

Word-to-Excel conversion typically takes under 10 minutes for most surveys. For example, converting 3 surveys with 106 questions and 490 options takes less than 10 minutes. Platform conversions (Excel to platform format) usually complete in 30–60 seconds.

No. Pirai AI is GDPR compliant by design and does not store your uploaded documents. When you upload a Word or PDF file, it stays in your browser's memory during processing and is automatically destroyed after conversion. Files are never saved to permanent storage. All processing happens in real-time with temporary handling and automatic cleanup. The platform uses privacy-first architecture. Only anonymized analytics (no personal or survey content) are logged for platform improvements.

Yes! Pirai AI uses enterprise-grade UTF-8 compliant processing to ensure perfect character encoding across all languages and platforms. Special characters, accents, and non-Latin scripts are preserved accurately throughout the conversion process.

Pirai AI uses a streamlined 2-step workflow: Step 1 (Document to Excel) — Upload your Word (.docx) or PDF survey document. Pirai AI's AI engine extracts all questions, answer options, and structure, then converts everything into the standardized 7-column Excel template. Review the typo check report and download your Excel file. Step 2 (Excel to Platform) — Upload your Excel file and select your target platform from 20+ options. Pirai AI converts the Excel template to the platform-specific format. Download the generated file with platform-specific import instructions. Total time: Under 10 minutes for Step 1, 30–60 seconds for Step 2.

Pirai AI uses a 2-step approach where Excel serves as the universal intermediate format. This design gives you maximum flexibility — you can review, edit, and validate your survey structure in Excel before converting to any platform. It also allows you to convert the same content to multiple platforms without starting over.

Each platform converter in Pirai AI shows key capabilities like matrix question support, recode value handling, and complexity rating. The platform displays whether your specific survey features (like matrix grids or custom scoring) are supported natively or will be converted to alternative formats. This helps you choose the platform that best matches your survey requirements.

Each platform converter in Pirai AI shows key capabilities like matrix question support, recode value handling, and complexity rating. The platform displays whether your specific survey features are supported natively or will be converted to alternative formats. This helps you choose the platform that best matches your survey requirements.

Pirai AI intelligently converts unsupported features to the closest equivalent. For example, matrix questions become individual single-choice questions on platforms without matrix support, and ranking questions may convert to multiple choice. You always receive clear warnings about any conversions made, so you know exactly how your survey will appear in the target platform.

No question limit! Pirai AI handles surveys of any size, from 10 questions to 200+ questions. The intelligent chunking system automatically processes large documents while maintaining accuracy and completeness throughout the entire survey.

Pirai AI accepts Word and PDF files up to 5MB, which accommodates surveys with hundreds of questions. If you have an exceptionally large document, consider splitting it into logical sections or reach out to admin@piraiai.com for assistance with enterprise-scale conversions.

Pirai AI uses advanced language models optimized for survey structure recognition with low-temperature settings for consistent, accurate results. The system includes automatic quality validation and spell-checking. While the AI is highly accurate, we recommend reviewing the Excel output before Step 2 — that's why we make it easy to download, edit, and re-upload.

Yes! Pirai AI is perfect for migrating surveys between platforms. If you have an existing survey specification, convert it to Excel in Step 1, then convert to your new target platform in Step 2. This eliminates the need to manually rebuild surveys when switching tools or managing multi-platform deployments.

Yes. After Step 1, you can share the Excel file with team members for review and editing. Multiple people can review questions, fix typos, or adjust wording in Excel. Once finalized, anyone can upload the edited Excel file in Step 2 to generate platform files. This makes Pirai AI ideal for research teams, agencies, and organizations with approval workflows.

Absolutely! Pirai AI supports research-focused platforms like Qualtrics, QuestionPro, REDCap, and Forsta. The system preserves complex question structures, matrix grids, and custom recode values essential for research data analysis. The multilingual support and privacy-first design make it ideal for international and sensitive research projects.

Yes! Agencies and consultants use Pirai AI to accelerate survey programming for clients. Convert client specifications to multiple platform formats quickly, handle multi-regional deployments with multilingual support, and deliver projects faster while maintaining quality. Contact admin@piraiai.com for enterprise pricing and agency support.

Manual survey programming typically takes hours or days and is error-prone, especially for large surveys. Pirai AI converts the same survey in under 10 minutes with consistent accuracy. The AI handles tedious data entry, formatting, and structure conversion automatically, while you focus on content quality and platform-specific optimizations.

Qualtrics Advanced format (.txt) is a human-readable text format you import through the "Import from Word" feature — great for reviewing before import. QSF format (.qsf) is Qualtrics' native JSON format that you import through "Import Survey" — it creates a complete, ready-to-use survey instantly. Both formats work perfectly; choose based on your preference and workflow.

The Excel template uses 7 standardized columns: Block Name, Question ID, Question Text, Question Type, Answer Option Text, Recode Value, and Matrix Type. This structure is designed to capture all survey elements in a platform-agnostic format that can be converted to any target platform.

Yes, for platforms that support custom recode values (like Qualtrics, REDCap, Forsta). The Recode Value column in the Excel template preserves your custom coding scheme. For platforms without recode support, the system uses simplified sequential numbering.

The current version focuses on question and response conversion. Advanced features like display logic, skip logic, and branching are handled through the platform's native tools after import. Pirai AI ensures the foundational survey structure is perfect for you to add platform-specific logic.

Each platform conversion includes detailed import instructions specific to that platform. For example, Qualtrics QSF files can be imported directly through Survey → Create New → Import Survey. Instructions are tailored to the file format and platform's import process.

This usually happens if the document is corrupted, password-protected, or has an unsupported format. Ensure your file is a standard .docx or PDF, is not password-protected, and is under 5MB. If the problem persists, contact admin@piraiai.com.

Error messages tell you what went wrong — like "duplicate Question IDs" or "matrix ordering issues." Click the Smart Fix button to auto-correct most errors, or read the specific error description to understand what needs fixing. If unclear, copy the error message and ask AI Editor "To fix the error."

Your browser might be blocking the download, or validation failed with critical errors. Try a different browser (Chrome or Firefox work best) if the problem persists.

The preview might still be loading, or there's a browser compatibility issue. Wait 5–10 seconds for the preview to render, then refresh the page if nothing appears. If using Safari, try Chrome or Firefox instead — they have better compatibility with Pirai AI's preview features.

Your request might be too complex, or the session timed out. Try breaking your request into smaller commands (e.g., "Add option to Q3" instead of editing multiple questions at once). If AI Editor stops responding completely, refresh the page — your survey data is saved and you can continue editing.

If you're in the same browser session, your survey is temporarily saved and you can refresh the page to recover it.

Yes! Email admin@piraiai.com with your question. Include screenshots and describe what happened — the more details you provide, the faster we can help. Check the FAQ first as most common issues have quick solutions listed here.

Smart Fix is a one-click AI tool that automatically corrects structural errors in your Excel template after document conversion. It fixes technical formatting issues like matrix ordering, duplicate Question IDs, and empty required fields without changing your survey content. Think of it as an auto-correct for survey structure.

Use Smart Fix immediately after Quality Check shows validation errors. It's most helpful when you have multiple structural errors (5+ issues) that would take time to fix manually. Click the Smart Fix button right after reviewing your validation errors.

Smart Fix corrects structural formatting: reorders matrix rows (Statements before Scale Points), removes duplicate Question IDs, empties Question Text in option rows, fixes sequential numbering gaps, and removes programming instructions leaked into questions. It fixes the Excel structure, not the question wording or content.

Smart Fix doesn't change question wording, answer options, survey content, or typos — it only fixes technical Excel structure. It won't add missing questions from your original document, fix character encoding corruption, or improve question clarity. For content changes, use AI Editor instead.

No, Smart Fix only fixes structural errors and never deletes content or changes question meaning. All changes are tracked in version history, so you can undo if needed. The worst case is Smart Fix doesn't fully resolve all errors — you'd then use AI Editor for remaining issues.

If Smart Fix reduced errors but some remain, use AI Editor to fix specific issues by typing commands like "Fix matrix Q5 ordering." If Smart Fix made no changes, your errors might be content-related (typos, missing text) rather than structural. Check the remaining error messages and use AI Editor with "Include original document" for context-aware fixes.

Only use Smart Fix when Quality Check shows validation errors. Most simple documents (10–30 questions) convert cleanly without needing Smart Fix. Complex documents (50+ questions, many matrices) usually benefit from Smart Fix.

Yes! Click the ↩️ Undo button or type "Undo" in AI Editor to roll back Smart Fix changes instantly. Every Smart Fix operation creates a version snapshot in your history. If you undo and still need fixes, you can manually edit or use AI Editor for targeted corrections.

AI Editor is a conversational assistant that lets you edit surveys using natural language commands instead of manually editing Excel. Just type what you want changed (like "Add option 'Other' to Q5" or "Delete Q7"), and AI makes the changes instantly. It's like having a survey programmer who understands plain English.

Type your request in plain English in the chat box and click "Send ➤" or press Enter. Be specific: say "Change Q3 to multiple choice" instead of vague requests like "fix Q3." Use simple commands like "Add," "Delete," "Change," or ask questions like "What platforms support matrix questions?"

Ask AI Editor to make edits ("Add option to Q5"), get survey advice ("Should demographics go at the end?"), answer questions about Pirai AI ("Which platforms support matrices?"), or request recommendations ("Review my survey and suggest improvements"). It handles editing commands, consultation questions, and platform guidance all in one conversation.

Yes! Say "Add a new question after Q5: What is your income?" and AI Editor will insert it with proper formatting. You can specify question type, answer options, and position. AI Editor assigns the correct Question ID, updates numbering, and maintains Excel structure automatically.

Yes! Type "Delete Q7" and AI Editor removes the entire question including all its answer options and rows. It automatically renumbers subsequent questions (Q8 becomes Q7, Q9 becomes Q8, etc.) to maintain sequential order. You can undo if you deleted the wrong question.

Yes! Say "Change Q3 from single choice to multiple choice" or "Convert Q5 to matrix question" and AI Editor updates the Question Type column. For complex conversions (like text entry to matrix), provide the new structure details. AI Editor adjusts all related columns to match the new question type.

Yes! Type "Fix spelling errors in Q3" or "Change 'recieve' to 'receive' in Q7" and AI Editor corrects them. For multiple typos, say "Fix all spelling errors" and AI Editor scans and corrects them. It can also fix grammar, punctuation, and formatting issues in question text or answer options.

AI Editor sees your current Excel state including all previous edits made in this session, but doesn't remember conversations across different sessions or surveys. Each version snapshot shows cumulative changes. If you refresh the page or start a new conversion, AI Editor starts fresh with no memory of past edits.

Click the ↩️ Undo button or type "Undo" to instantly roll back the last change. AI Editor saves every change as a version, so you can undo multiple times. If the edit is partially wrong, ask AI Editor to "Fix Q5 by..." with more specific instructions rather than undoing.

Yes! Every AI Editor change creates a version snapshot. Click ↩️ Undo or type "Undo" to roll back the last edit. You can undo multiple times to go back several versions. Version history shows all changes with timestamps, so you can see exactly what was modified.

Unlimited! Make as many edits as needed in a single session — add questions, delete options, fix typos, change types, ask questions. There's no limit on requests per survey or per session. Each edit is tracked in version history, and you can undo/redo freely.

Version history automatically saves a snapshot of your survey after every change made by Smart Fix or AI Editor. Each version is timestamped and stored temporarily during your session, allowing you to roll back mistakes or compare different versions. Think of it as "Track Changes" for survey editing.

Click the ↩️ Undo button in the interface, or type "Undo" in AI Editor and press Enter. This instantly rolls back your survey to the previous version before the last change. The undo happens immediately — you'll see the preview update to show the previous state.

All versions created during your current session are saved — there's no limit on version count. If you make 20 edits with AI Editor or Smart Fix, you'll have 20 versions in history.

Yes! Click ↩️ Undo multiple times to step backward through your version history. Each undo takes you one version back — undo twice to go back two changes, three times for three changes, etc. You can keep undoing until you reach your original converted survey.

Validation errors are structural issues that prevent your Excel from importing correctly into survey platforms. They're technical formatting problems (like matrix rows in wrong order or duplicate Question IDs), not problems with your survey content.

Errors happen because AI converts unstructured documents (Word/PDF) into highly structured Excel format, and complex layouts sometimes confuse the AI. Documents with nested tables, mixed formatting, or programming instructions often produce structural errors. This is normal — that's why Smart Fix exists to auto-correct them.

Errors (🔴 HIGH severity) prevent platform import and must be fixed. Warnings (🟡 MEDIUM, 🟢 LOW severity) are suggestions for improvement but won't stop import — you can download with warnings. Fix HIGH first, MEDIUM if time permits, LOW is optional.

Yes! Warnings (🟡 MEDIUM and 🟢 LOW) won't prevent platform import — they're suggestions to improve quality or consistency. Only 🔴 HIGH severity errors are mandatory to fix.

Structural errors are Excel formatting issues: matrix Scale Points appearing before Statements, duplicate Question IDs, empty Question Text in the first row, non-sequential numbering (Q1, Q2, Q4 skips Q3), or Question Text repeating in option rows. These break the Excel template rules that platforms expect.

Content errors are problems with survey text: spelling mistakes, character encoding issues (replacement character symbols in multilingual text), programming instructions leaked into questions ("ASK IF Q1=2"), or incomplete question wording. Smart Fix doesn't handle these — use AI Editor or manual correction instead.

AI sometimes reuses the same Question ID (like "Q5") for two different questions during conversion, especially in complex documents with nested questions or unclear numbering. This confuses platforms because each question needs a unique ID. Smart Fix automatically renumbers to make all Question IDs unique.

Your original document might have duplicate Question IDs, and AI tries to preserve your original numbering whenever possible. If the document has unclear or no numbering, AI assigns IDs but may accidentally reuse the same ID for different questions. This happens especially in complex documents with nested questions or sub-questions (Q5a, Q5b).

Yes! AI attempts to follow the Question IDs in your original document first (Q1, Q2, Q3 or 1., 2., 3.). If your document has clear numbering, AI preserves it exactly as written. Only when numbering is unclear, missing, or ambiguous does AI assign new sequential IDs automatically.

AI will convert the document as-is, preserving the duplicate IDs from your original, which triggers validation errors. You'll need to fix the duplicates — either use Smart Fix to auto-renumber all questions sequentially, or use AI Editor to manually assign specific IDs. The validation report will show exactly which Question IDs are duplicated.

Yes! Type "Renumber all questions sequentially starting from Q1" and AI Editor will assign Q1, Q2, Q3… in order, removing all duplicates. You can also specify custom numbering like "Renumber questions starting from Q10" or "Renumber Block 2 questions as Q20, Q21, Q22." AI Editor updates all rows for each question automatically.

Use Smart Fix first — it automatically renumbers to remove duplicates in seconds. If Smart Fix doesn't fully resolve it (like when you want specific numbering patterns), then use AI Editor with commands like "Change the second Q5 to Q6 and renumber subsequent questions." Manual Excel editing works too but is slower.

You have three options: fix all errors using Smart Fix or AI Editor and download the complete survey; download only error-free questions using "Download Clean File" to proceed immediately; or fix errors manually in Excel after downloading. Most users try Smart Fix first since it resolves 80–90% of errors automatically in seconds.

Yes! Use "Skip & Continue" to export only validated questions and skip problematic ones temporarily. This lets you start building your survey in the target platform right away with working questions. You can fix error questions separately later and add them to your survey, or merge them using Step 2's multi-file feature.

Click the "Skip & Continue" button (appears when validation shows errors) to export an Excel containing only questions that passed all validation checks. The download excludes any rows flagged with 🔴 HIGH severity errors. You'll get a clean, platform-ready Excel file with your validated questions immediately.

Yes! The "Skip & Continue" option exports an Excel with only questions that passed validation, automatically excluding any rows with errors. This is useful when you have a few problematic questions but want to proceed with the rest of your survey immediately.

Error questions are excluded from the "Skip & Continue" export — they won't appear in the downloaded Excel. Your original survey with all questions (including errors) remains in the browser session, so you can continue editing. The error questions aren't deleted permanently; they're just filtered out of the clean download.

Fix all errors first using Smart Fix or AI Editor, then click the standard Download button to get a complete Excel with all questions. Alternatively, use Smart Fix to resolve most structural errors automatically, download, then manually fix any remaining content issues in Excel. Most structural errors take seconds to fix with Smart Fix.

Yes! Fix the error questions in your original session (using Smart Fix or AI Editor), download them as a separate Excel file, then use Step 2's multi-file merge feature to combine them with your previously converted survey. This lets you build and deploy surveys incrementally without losing work.

Use AI Editor to delete the questions you don't need: "Delete Q7" removes it from the Excel. Alternatively, download the full Excel and manually delete rows before uploading to Step 2. The "Skip & Continue" feature specifically handles questions with validation errors, not intentional omissions.

Yes, Step 1 always downloads as Excel (.xlsx) — the universal intermediate format for the Pirai AI workflow. Step 2 converts this Excel to your chosen platform format (e.g., .qsf for Qualtrics, .zip for REDCap, .csv for SurveyMonkey). Each platform has its own output format with specific import instructions provided.

Qualtrics Advanced (TXT format) supports ALL question types natively: MC:SingleAnswer, MC:MultipleAnswer, Matrix, TE:Essay, TE:SingleLine, TE:Form, DB (Descriptive Text), Page Break, Rank Order, and Constant Sum. No conversions needed — questions import exactly as designed with full formatting, logic, and display capabilities. Qualtrics Advanced is one of the most feature-complete platforms, ideal for complex research surveys.

Qualtrics QSF (JSON format) supports ALL question types natively with instant survey creation: MC:SingleAnswer, MC:MultipleAnswer, Matrix, TE:Essay, TE:SingleLine, TE:Form, DB, Page Break, Rank Order, and Constant Sum. QSF creates a complete, ready-to-use survey in one import with no manual setup required. This is the fastest way to get surveys into Qualtrics.

Native support: MC:SingleAnswer (radio buttons), MC:MultipleAnswer (checkboxes), Matrix (converted to individual radio questions per statement), TE:Essay (notes field), TE:SingleLine (text field), TE:Form (multiple text fields), and DB (descriptive text). Converted with limitations: RO:RankOrder becomes a single-choice radio button (loses ranking functionality); CS:ConstantSum becomes a text input (loses validation). Not supported: Page Break — REDCap uses a different section/event structure for pagination. REDCap is optimised for clinical research with strong validation and compliance features.

Native support: MC:SingleAnswer, TE:Essay, TE:SingleLine. Converted: MC:MultipleAnswer → single-select (changeable manually after import); TE:Form → multiple individual textbox questions; Matrix → individual textbox questions (statements only, scale points ignored); RO:RankOrder and CS:ConstantSum → multiple choice (lose specialised functionality). Not supported / skipped: DB (Descriptive Text) and Page Break. SurveyMonkey is a consumer-friendly platform best suited to straightforward surveys.

Native support: MC:SingleAnswer, MC:MultipleAnswer, Matrix, TE:Essay, TE:SingleLine, TE:Form. Converted: RO:RankOrder → MC:SingleAnswer (loses ranking functionality). Not supported / skipped: DB, Page Break, CS:ConstantSum. QuestionPro is excellent for market research but requires questions only — no descriptive blocks.

Native support: MC:SingleAnswer, MC:MultipleAnswer, Matrix, TE:Essay, TE:SingleLine, DB, RO:RankOrder, CS:ConstantSum. Converted: TE:Form — each form field becomes a separate text_single_row question via individual API calls. Not supported: Page Break (QuestionPro handles pagination differently). The API format provides programmatic control, real-time deployment, comprehensive response logging, and automatic retry logic.

Native support: MC:SingleAnswer, MC:MultipleAnswer, TE:Essay, TE:SingleLine. Converted: Matrix → individual single-select questions (one per statement, scale points become answer options); TE:Form → multiple individual text questions; RO:RankOrder → single-select (loses ranking); CS:ConstantSum → text input (loses validation). Not supported / skipped: DB and Page Break. Forsta (Word) is a market research platform focused on questions rather than descriptive content.

Forsta Decipher supports ALL question types with native scripting format: MC:SingleAnswer, MC:MultipleAnswer, Matrix, TE:Essay, TE:SingleLine, TE:Form, DB, Page Break, Rank Order, and Constant Sum. Decipher generates survey programming script (not GUI import), giving maximum control over question logic and display. Best for advanced survey programmers needing full flexibility.

Native support: MC:SingleAnswer ([radio]), MC:MultipleAnswer ([checkbox]), Matrix ([cols]/[rows]), TE:Essay ([comment]), TE:SingleLine ([text]), Page Break. Converted: TE:Form → multiple individual [text] questions; RO:RankOrder → [radio] single-choice (loses ranking); CS:ConstantSum → [text] (loses validation). Not supported / skipped: DB. BConvert is optimised for market research professionals needing Confirmit-style programmatic import.

Native support: MC:SingleAnswer, MC:MultipleAnswer, Matrix, TE:Essay, TE:SingleLine, DB, Page Break. Converted: TE:Form → separate text entry questions (one per field); RO:RankOrder → categorical single-choice (loses ranking); CS:ConstantSum → categorical single-choice (loses sum validation). MDD is IBM SPSS Data Collection format for enterprise survey data management with loop structures for matrices and info items for descriptive blocks — excellent for large-scale research requiring sophisticated data analysis.

Native support: MC:SingleAnswer, MC:MultipleAnswer, Matrix (as grid), TE:Essay, TE:SingleLine, DB (as section headers). Converted: RO:RankOrder → multiple choice (ranking functionality lost). Not supported / skipped: TE:Form, CS:ConstantSum, Page Break. Google Forms is simple and free but has limitations for complex surveys.

Native support: MC:SingleAnswer, TE:Essay, TE:SingleLine. Converted: MC:MultipleAnswer → single choice (Microsoft Forms limitation); Matrix → individual single-choice questions (one per statement); TE:Form → individual text questions (one per field); RankOrder → multiple choice; ConstantSum → text question. Not supported / skipped: DB and Page Break. Microsoft Forms is user-friendly but limited for complex question types, descriptive blocks, or page controls.

Native support: MC:SingleAnswer, TE:Essay, TE:SingleLine, Matrix. Converted to single-answer format: MC:MultipleAnswer, TE:Form, RO:RankOrder, CS:ConstantSum. Not supported / skipped: DB and Page Break. Zoho is good for business surveys with CRM integration but lacks descriptive block support and converts multiple answer types to single choice.

Native support: MC:SingleAnswer, TE:Essay, TE:SingleLine. Converted to single-answer format: MC:MultipleAnswer, Matrix (one question per statement), TE:Form, RO:RankOrder, CS:ConstantSum. Not supported / skipped: DB and Page Break. SurveySparrow is known for its conversational survey interface and high completion rates but doesn't support descriptive blocks or native matrix/grid questions.

Native support: MC:SingleAnswer only. Converted: MC:MultipleAnswer → single-answer (loses multi-select functionality — Aiken format limitation). Not supported / skipped: Matrix, TE:Essay, TE:SingleLine, TE:Form, RO:RankOrder, CS:ConstantSum, DB, Page Break. Moodle (Aiken format) is education-focused for simple single-choice quizzes and exams only. For advanced question types including matrix and text questions, use Moodle XML format instead.

Native support: MC:SingleAnswer only for quiz creation. Not supported / skipped: MC:MultipleAnswer, Matrix, TE:Essay, TE:SingleLine, TE:Form, RO:RankOrder, CS:ConstantSum, DB, Page Break. Canvas (Word format) is an LMS focused on simple assessments, not full surveys.

Native support: MC:SingleAnswer (quiz format). Converted: MC:MultipleAnswer → single-answer (loses multi-select functionality). Not supported / skipped: Matrix, TE:Essay, TE:SingleLine, TE:Form, RO:RankOrder, CS:ConstantSum, DB, Page Break. Kahoot is designed for interactive, game-based learning and quick polls. Additional constraints: maximum 4 answer options per question; question length 2–120 characters.

Native support: MC:SingleAnswer, MC:MultipleAnswer, TE:SingleLine. Converted: TE:Essay → Short Text (TST); TE:Form → multiple Short Text questions (one per field); Matrix → multiple single-choice questions (TMC, one per statement); RO:RankOrder → Dropdown (TDD); CS:ConstantSum → Short Text (TST). Not supported / skipped: DB and Page Break. LearnWorlds is an e-learning platform focused on course delivery with assessment capabilities.

Native support: MC:SingleAnswer, MC:MultipleAnswer, TE:SingleLine, TE:Essay. Converted: TE:Form → free text; Matrix → multiple single-choice questions (one per statement); RO:RankOrder → single-choice; CS:ConstantSum → free text. Not supported / skipped: DB and Page Break. BrainCert is a unified training platform for education and corporate training.

Native support: MC:SingleAnswer and MC:MultipleAnswer only for compliance training and assessments. Not supported / skipped: Matrix, TE:Essay, TE:SingleLine, TE:Form, RO:RankOrder, CS:ConstantSum, DB, Page Break. Atomic focuses on workplace training and certification with simple multiple-choice testing. Minimum requirement: 2 answer options per question.

Native support: MC:SingleAnswer. Converted: MC:MultipleAnswer → single choice (multiple selection lost — must be adjusted manually in ClassMarker after import). Not supported / skipped: Matrix, TE:Essay, TE:SingleLine, TE:Form, RO:RankOrder, CS:ConstantSum, DB, Page Break. ClassMarker is designed for secure online exams. Constraints: maximum 50 questions per CSV file; 2–10 answer options per question.

Native support: MC:SingleAnswer, MC:MultipleAnswer, TE:Essay, TE:SingleLine. Converted: TE:Form → multiple fill-in-the-blank questions (one per form field). Not supported / skipped: Matrix, RO:RankOrder, CS:ConstantSum, DB, Page Break. Respondus is a quiz authoring tool for importing assessments into LMS platforms (Blackboard, Canvas, Moodle) with support for multiple choice, multiple response, essay, and fill-in-the-blank types.

QTI (Question & Test Interoperability) supports ALL question types: MC:SingleAnswer, MC:MultipleAnswer, Matrix, TE:Essay, TE:SingleLine, TE:Form, DB, Page Break, RO:RankOrder, and CS:ConstantSum. QTI is an IMS Global standard for e-learning assessment exchange with comprehensive question type coverage. Full support depends on the specific QTI implementation version and target LMS (Canvas, Blackboard, Brightspace, Moodle).

Native support: MC:SingleAnswer, MC:MultipleAnswer, Matrix, TE:Essay, TE:SingleLine. Converted: RO:RankOrder → radio button single choice (ranking functionality lost). Not supported / skipped: TE:Form, DB, Page Break, CS:ConstantSum. Alchemer (formerly SurveyGizmo) is enterprise-focused with advanced logic but lacks support for descriptive blocks, form fields, and constant sum questions.

Imported as "Text Choice" by default: MC:SingleAnswer, MC:MultipleAnswer, RO:RankOrder — users must manually change question types to Multiple Choice, Rating, etc. after import. Ranking questions lose ranking functionality and become single-choice. Not supported / skipped: Matrix, TE:Essay, TE:SingleLine, TE:Form, CS:ConstantSum, DB, Page Break. PointerPro specialises in scored assessments and recommendation engines for consultative selling.

CSV UTF-8 is a universal translation export format that preserves ALL question types in simple tabular structure: MC:SingleAnswer, MC:MultipleAnswer, Matrix, TE:Essay, TE:SingleLine, TE:Form, DB, RankOrder, and ConstantSum. Not included: Page Break. This format outputs QuestionID and Text columns for multilingual translation workflows — export from bilingual documents or send to translation agencies. Not intended for platform import.

Qualtrics Advanced exports as .txt (text/plain). The file contains survey questions in Qualtrics Advanced Format with special syntax for question types, display logic, and answer choices. Import via Qualtrics → Survey → Tools → Import/Export → Import Survey → Advanced Format.

Qualtrics QSF exports as .qsf (application/json). The QSF (Qualtrics Survey Format) is a JSON-based format containing complete survey structure including questions, logic, display settings, and metadata. Import via Qualtrics → Create Project → Import QSF file.

REDCap exports as .csv (text/csv). The file contains data dictionary format with question metadata in CSV structure. Import via REDCap → Project Setup → Data Dictionary → Upload Data Dictionary.

SurveyMonkey exports as .txt (text/plain). The file contains questions and answer choices formatted for SurveyMonkey's paste import method. Import instructions: (1) Click Create Survey in SurveyMonkey. (2) Select "Paste your content". (3) On the Import Questions screen, copy and paste your Multiple Choice or Single Textbox questions and answer choices from the downloaded .txt file. Each question and answer choice must be on their own line. (4) Press Enter or Return twice to separate questions — each question must be separated by a blank line. (5) Preview section shows questions in real time as you paste. (6) Click "Add Questions" to create your survey. Note: Only Multiple Choice and Single Textbox question types are supported via paste import.

QuestionPro (Word) exports as .docx (application/vnd.openxmlformats-officedocument.wordprocessingml.document). The file contains formatted survey questions in Word document format. Import via QuestionPro → Survey → Tools → Import Survey → Upload Word Document.

QuestionPro API exports as .json (application/json). The file contains survey structure in JSON format for direct API import. Use QuestionPro API endpoints to import the survey programmatically.

Forsta (Word) exports as .docx (application/vnd.openxmlformats-officedocument.wordprocessingml.document). The file contains survey questions in Word document format compatible with Forsta's import system.

Forsta Decipher exports as .xml (application/xml). The file contains complete survey structure in Decipher XML format with full question logic and display settings. Import via Decipher survey editor → Import XML.

Forsta BConvert exports as .txt (text/plain). The file contains survey syntax in BConvert format for legacy Forsta survey systems. Import via BConvert tool.

Dimensions MDD exports as .mdd (application/octet-stream). The file contains Metadata Document Definition in binary format for SPSS Dimensions. Import via Dimensions Data Model editor.

Google Forms exports as .txt (text/plain). The file contains Google Apps Script code that programmatically creates your survey with all questions, answer options, and settings. Import via Google Forms → Click the three dots (⋮) → Script editor → Paste the code → Run the script. The script automatically builds your complete form.

Microsoft Forms exports as .docx (application/vnd.openxmlformats-officedocument.wordprocessingml.document). The file contains formatted questions for Microsoft Forms Quick Import. Import via Microsoft Forms → Quick Import → Upload Word document.

Zoho Surveys exports as .txt (text/plain). The file contains questions and answer choices in text format for bulk import. Import via Zoho Surveys → Create Survey → Add Questions in Bulk → Paste Text.

SurveySparrow exports as .txt (text/plain). The file contains survey questions in text format for bulk question import. Import via SurveySparrow → Create Survey → Add Questions in Bulk → Paste Text.

Moodle exports as .txt (text/plain). The file contains questions in Aiken format for quiz import. Import via Moodle → Question Bank → Import → Aiken format.

Canvas exports as .docx (application/vnd.openxmlformats-officedocument.wordprocessingml.document). The file contains quiz questions in Word format for Canvas LMS import. Import via Canvas → Quizzes → Question Banks → Import Questions → Word Document.

Kahoot exports as .xlsx (application/vnd.openxmlformats-officedocument.spreadsheetml.sheet). The file contains quiz questions in Excel spreadsheet format. Import via Kahoot → Create → Add question → Import spreadsheet.

LearnWorlds exports as .xlsx (application/vnd.openxmlformats-officedocument.spreadsheetml.sheet). The file contains assessment questions in Excel format for LearnWorlds Question Bank. Import via LearnWorlds → Assessment → Add question → Import Questions.

BrainCert exports as .csv (text/csv). The file contains test questions in CSV format with 16 columns for question metadata. Import via BrainCert → Assessments → Questions → Import from CSV.

Atomic Assessments exports as .csv (text/csv). The file contains questions in CSV format with extensive metadata columns. Import via Atomic Assessments → Content Manager → Add Content → Upload CSV.

ClassMarker exports as .csv (text/csv). The file contains multiple choice questions in CSV format. Maximum 50 questions per file. Import via ClassMarker → Tests → Question Bank → Import Questions → Upload CSV.

Respondus exports as .docx (application/vnd.openxmlformats-officedocument.wordprocessingml.document). The file contains exam questions in Respondus Word format with special type declarations (Type: MC, Type: MA, Type: E, Type: F). Import via Respondus → Import Questions → Word Document.

QTI exports as .zip (application/zip). The package contains QTI 2.1 XML files and manifest for universal LMS import. Import via your LMS: Canvas (Settings→Import), Blackboard (Tests→Import), Brightspace (Import/Export), Moodle (Question Bank→Import→QTI 2.1).

Alchemer exports as .docx (application/vnd.openxmlformats-officedocument.wordprocessingml.document). The file contains survey questions in Word format for Alchemer import. Import via Alchemer → Create Survey → Import from Word → Paste Content.

PointerPro exports as .txt (text/plain). The file contains questions in text format where all questions import as "Text Choice" by default. Import via PointerPro → Create Survey → Import Questions → Paste Text. Manually adjust question types after import.

CSV UTF-8 exports as .csv (text/csv). The file contains QuestionID and Text columns in UTF-8 encoding with BOM for Excel compatibility. Designed for translation workflows — export from bilingual docs or send to translation agencies. Not intended for platform import.

Pirai AI accepts Word documents (.docx), PDF files (.pdf), plain text (.txt), Excel spreadsheets (.xlsx), CSV files (.csv), Qualtrics QSF (.qsf), and XML files (.xml). For best results, use Word (.docx) or PDF (.pdf) formats. Convert Google Docs to Word or PDF before uploading.

Yes! Use the "PDF Advanced" extraction method for scanned PDFs — it uses Vision AI to read the document like a human would. Standard extraction won't work on scanned PDFs because they're images, not text. PDF Advanced takes longer (30–60 seconds per page) but extracts content that standard methods cannot.

Yes! Convert the photo to PDF first, then upload using "PDF Advanced" extraction method. Vision AI can read printed text from photos, but ensure the image is clear, well-lit, and not too blurry. Higher quality photos (300+ DPI) produce better extraction accuracy.

Yes! Upload .xlsx or .csv files directly. If your Excel is already in Pirai AI's 7-column format (Block Name, Question ID, Question Text, Question Type, Answer Option Text, Recode Value, Matrix Type), skip Step 1 and upload directly to Step 2. If not, Step 1 will restructure it to the universal template.

No, you must convert Google Docs to Word (.docx) or PDF (.pdf) first. In Google Docs, click File → Download → Microsoft Word (.docx) or PDF, then upload the downloaded file to Pirai AI. The conversion is instant and preserves formatting.

The file size limit is 5MB for most formats. Documents over 5MB should be split into smaller files or compressed. Very large surveys (200+ questions) work fine as long as file size stays under 5MB. PDFs with many high-resolution images may exceed the limit.

Common reasons: file is password-protected (unlock it first), file is corrupted (try saving a fresh copy), unsupported format (convert to .docx or .pdf), or file exceeds 5MB size limit (compress or split). If using an older Word format (.doc), save as .docx first.

Your Excel workbook might contain multiple tabs: one with survey questions, others with data, notes, templates, or archived versions. Pirai AI processes all visible sheets, which can cause confusion or errors if extra sheets contain unrelated data. Keep only the sheet with your actual survey questions for best results.

Yes, strongly recommended! Delete all sheets except the one containing your survey questions to avoid processing errors and confusion. Extra sheets (data, notes, templates) can slow down conversion and may cause AI to mix unrelated content into your survey. Right-click sheet tabs and select "Delete" for unwanted sheets.

Hidden sheets are processed by Pirai AI just like visible sheets, which can cause unexpected content to appear in your conversion or trigger errors. Unhide all sheets first (Right-click any tab → Unhide → select sheet), review the content, then delete unwanted sheets before uploading. Only keep your survey questions sheet.

Pirai AI processes ALL sheets in your Excel workbook (both visible and hidden), combining content from every sheet into the conversion. If you have multiple sheets, AI treats them as one continuous survey. Delete all sheets except your survey questions sheet to ensure clean, accurate conversion.

Yes, Pirai AI automatically processes every sheet in your workbook and combines them into one survey. However, this causes problems if sheets contain different data (notes, calculations, templates). For best results, keep only ONE sheet with your survey questions and delete all others before uploading.

Yes! Keep only the sheet containing your survey questions and delete all other sheets (data tabs, notes, templates, archived versions). Having one clean sheet ensures AI processes exactly what you need without confusion. Right-click unwanted sheet tabs, select "Delete," then upload the cleaned Excel file.

Pirai AI processes all sheets and combines their content into one survey, which can mix survey questions with unrelated data, notes, or templates. This causes validation errors, incorrect question counts, or garbled content. Always delete extra sheets before uploading to ensure only your survey questions are processed.

In Excel: Right-click the sheet tab you want to remove → Click "Delete" → Confirm. Repeat for all unwanted sheets until only your survey questions sheet remains. In Google Sheets: Click the small arrow on the sheet tab → "Delete" → Confirm. Save the file before uploading to Pirai AI.

Yes! Hidden sheets are processed just like visible sheets, which can inject unwanted content into your survey or cause errors. Unhide all sheets first (Right-click any sheet tab → "Unhide" → select hidden sheet), review what's there, delete unwanted sheets, then upload. Hidden sheets are a common cause of unexpected conversion results.

Click the 🔧 Smart Fix button immediately after seeing validation errors — it auto-corrects 80–90% of structural errors in seconds. Review the validation report to see which errors Smart Fix resolved, then use AI Editor for any remaining issues. Start with Smart Fix before manual editing because it's faster and handles most common problems automatically.

Yes! Download the Excel file, fix errors directly (reorder matrix rows, remove duplicate IDs, etc.), save, then re-upload to Step 2 for platform conversion. Manual fixing gives you complete control but takes longer than Smart Fix. This approach works best if you're comfortable with Excel and prefer hands-on editing.

Use AI Editor for remaining errors — check "Include original document for AI reference" and type specific commands like "Fix matrix Q5 ordering" or "Remove duplicate Question IDs." If AI Editor can't resolve complex issues, download Excel, fix manually, and re-upload. Some errors (like missing content) require human judgment that Smart Fix cannot provide.

Fix all 🔴 HIGH severity errors first — these prevent platform import and must be resolved. Then fix 🟡 MEDIUM warnings if time permits (improve quality but not critical). Ignore 🟢 LOW warnings unless you want perfect scores. Use Smart Fix to clear HIGH errors quickly, then prioritize remaining issues by severity color.

Yes! Click "Download Clean File" to export only validated questions, excluding any with errors. This lets you proceed with working questions immediately while fixing problematic ones separately later. You can merge the fixed questions back using Step 2's multi-file feature once errors are resolved.

Try re-uploading with "PDF Advanced" extraction (converts to PDF first if needed) — Vision AI handles multilingual text better than standard extraction. Or use AI Editor with "Include original document" checked and say "Fix character encoding in Q3 — replace garbled text with correct [language] text from original." Manual copy-paste from original document also works.

Click Smart Fix — it automatically reorders matrix rows to put ALL Statements before ALL Scale Points. If Smart Fix doesn't work, use AI Editor: "Fix matrix Q5 — move all scale points after statements." Manual fix: download Excel, cut/paste rows to reorder (Statements first, then Scale Points), save, re-upload.

Yes! Pirai AI supports 50+ languages including English, Spanish, French, German, Chinese, Arabic, Hebrew, Japanese, Korean, and more. The platform uses UTF-8 encoding to preserve all character sets and special characters. For best results with multilingual surveys, use properly formatted Word documents or convert to PDF before uploading.

Yes, Pirai AI processes surveys in any language — Spanish, French, Arabic, Chinese, Hebrew, Japanese, etc. The AI understands question structures regardless of language and preserves all special characters and accents. Validation and Quality Check work identically for all languages, detecting structural errors while preserving your original text.

Pirai AI supports all major languages: English, Spanish, French, German, Italian, Portuguese, Chinese (Simplified/Traditional), Japanese, Korean, Arabic, Hebrew, Russian, Polish, Dutch, Swedish, Norwegian, Danish, Finnish, Greek, Turkish, Hindi, Thai, Vietnamese, and 30+ more. Any language that uses standard character encoding (UTF-8) works perfectly.

Yes! Surveys with multiple languages in one document (like English questions with Spanish answer options, or bilingual surveys) work fine. Pirai AI preserves all languages exactly as written and maintains proper character encoding. Ensure your Word document is properly formatted with consistent styles for best multilingual results.

This is a character encoding issue (UTF-8 corruption) that happens during document extraction, especially with right-to-left languages or non-Latin scripts. Try saving your document as PDF first, then use the "PDF Advanced" extraction method — it handles encoding better. Or use AI Editor with "Include original document" to fix corrupted text by referencing the source.

First, ensure your Word document uses UTF-8 encoding (File → Save As → More options → Encoding: UTF-8). If errors remain after upload, try converting to PDF first and re-uploading. Use AI Editor with "Include original document" checked and say "Fix character encoding in Q3 — replace garbled text with correct Arabic/Chinese/Hebrew from original."

Yes! Pirai AI preserves Arabic, Hebrew, Persian, and Urdu text with correct right-to-left directionality. The Excel template maintains proper text direction and character encoding. Use properly formatted Word documents with consistent RTL formatting, or convert to PDF if you experience text scrambling or encoding issues.

No, Pirai AI uses UTF-8 encoding to preserve all special characters: accents (é, ñ, ü), currency symbols (€, ¥, ₹), mathematical notation (±, ≤, ∑), and non-Latin scripts (Arabic, Chinese, Cyrillic). If you see � symbols or garbled text, it's an encoding error — ensure your Word document uses UTF-8 encoding or convert to PDF.

Yes! If your language uses standard text characters (Latin, Cyrillic, Arabic, Hebrew, Chinese, Japanese, Korean, Thai, etc.), Pirai AI handles it perfectly. The platform is language-agnostic — it identifies question structures based on formatting patterns, not language. Survey quality and conversion accuracy are identical regardless of language.

The � symbol indicates UTF-8 character encoding corruption — your special characters or non-English text weren't decoded correctly during extraction. This happens with Arabic, Chinese, Hebrew, or accented characters. Fix: ensure your Word document uses UTF-8 encoding, or convert to PDF and re-upload, or use AI Editor to replace corrupted text by referencing your original document.

PDF Advanced can help with encoding issues because Vision AI "sees" text rather than extracting character codes. However, properly formatted Word documents with UTF-8 encoding work equally well for most languages. Use PDF Advanced if you see garbled text after trying standard extraction, or if your document has complex multilingual layouts.

A good document has clear question numbering (Q1, Q2, Q3), consistent formatting throughout, simple table structures, and standard question types (multiple choice, text entry, matrix). Clean Word documents with proper headings, no nested tables, and no programming instructions convert with 90–95% accuracy and minimal errors.

Bad documents have inconsistent formatting, mixed numbering systems (Q1, 1a, i., Question 2), heavily nested tables, merged cells, programming instructions mixed into questions, scanned pages with poor quality, or multiple columns. These create structural errors requiring Smart Fix or manual corrections after conversion.

AI may have missed questions if your document has unclear formatting, text boxes containing questions, questions embedded in images, very small font sizes, or complex nested structures. Check your original document — if questions are in tables within tables or floating text boxes, try converting to PDF and using "PDF Advanced" extraction.

Questions might be in images (AI can't read images with standard extraction), in text boxes that were skipped, on pages that failed to process, or formatted in ways AI didn't recognize as questions. Use "PDF Advanced" for scanned documents or images, or use AI Editor to manually add missing questions by referencing your original document.

Garbled text usually means character encoding issues with special characters or non-English languages, corrupted document file structure, or OCR errors from scanned documents. Try converting your Word document to PDF and re-uploading, or use "PDF Advanced" extraction method which handles encoding better through Vision AI.

Complex tables with merged cells, nested tables, or irregular column counts confuse AI during conversion. Matrix questions may have Statements and Scale Points in wrong order. Use Smart Fix first to correct matrix ordering, then AI Editor for specific table fixes, or try "PDF Advanced" which interprets tables visually.

Yes, but only clear, legible handwriting using "PDF Advanced" extraction (Vision AI). Very messy or cursive handwriting may not be fully captured. Printed questions with handwritten responses work best — the printed structure is captured while legible handwritten text is extracted. Typed documents always produce better accuracy than handwriting.

Large documents (50+ questions, 100+ pages), scanned PDFs with "PDF Advanced" (30–60 seconds per page), documents requiring intelligent chunking (100+ questions split into parts), or complex tables/images slow processing. Standard Word documents under 30 questions typically convert in 15–30 seconds; large surveys may take 5–10 minutes.

Yes! Pirai AI uses intelligent chunking to automatically split large surveys into manageable segments, process each part, then reassemble into one complete Excel file. Surveys with 100–200 questions convert successfully in 8–12 minutes. Documents over 300 questions should be split into multiple files and merged in Step 2.

Large documents (50+ questions) are automatically split into chunks using intelligent chunking — AI processes chunk 1, then chunk 2 with context overlap, then reassembles all parts into one complete survey. This prevents memory issues and ensures all questions are captured. You'll see progress indicators showing which chunk is being processed.

Smaller files (10–30 questions) process faster (15–30 seconds) and have fewer structural errors because there's less content to analyze and organize. Large files require chunking, take longer (5–10 minutes), and have higher chance of structural errors from complex formatting. However, Pirai AI handles both — small files are just simpler.

Yes! If your survey has 200+ questions, split it into 2–3 separate Word documents (e.g., Q1–Q70, Q71–140, Q141–200). Convert each part separately in Step 1, download the Excel files, then upload all Excel files to Step 2 and use multi-file merge to combine them into one complete survey.

Images with decorative content (logos, clipart) are skipped — no problem if questions are in text. If questions or answer options are embedded in images, use "PDF Advanced" extraction so Vision AI can read text from images. Scanned paper surveys (entire document is images) require "PDF Advanced" — standard extraction reads zero content.

Tables are automatically detected and converted to matrix questions or multiple choice options depending on structure. Simple tables convert perfectly; complex nested tables may need Smart Fix to correct matrix ordering. If tables are garbled, try "PDF Advanced" extraction which interprets table structure visually.

Yes, but results vary. Two-column layouts work reasonably well with Smart Detection. Complex multi-column layouts (3+ columns, magazine-style) should be converted to PDF first, then uploaded with "PDF Advanced" extraction — Vision AI understands visual reading order better than text extraction which may scramble columns.

Nested tables (tables within tables) are challenging but Pirai AI handles them through context-aware processing. Smart Detection processes each table level separately and maintains hierarchy. If results are garbled, try "PDF Advanced" which sees nested structure visually, or simplify your document by flattening nested tables before upload.

Only with "PDF Advanced" extraction — Vision AI reads text embedded in images, scanned documents, photos of paper surveys, and complex visual layouts. Standard extraction (Smart Detection, Visual Layout, Alternative Parser) only reads text content, not images. Convert your document to PDF and select "PDF Advanced" for image-based content.

Matrix conversion issues usually happen because AI mixed Statement and Scale Point order during extraction, or your platform doesn't support matrix natively. Use Smart Fix to reorder matrix rows (all Statements before all Scale Points), or check if your target platform supports matrices — platforms like Google Forms and SurveyMonkey convert matrices to individual MC questions automatically.

Option order issues occur when AI misinterprets complex table structures or nested formatting during document extraction. Check the Excel preview in Table View — if order is wrong there, use AI Editor to reorder: "Move option 3 to position 1 in Q5." If Excel is correct but platform output is wrong, it's a platform import issue (rare).

Missing question text means AI didn't extract it from your document (embedded in image, text box, or unclear formatting), or the Question Text cell is empty in Excel. Use AI Editor with "Include original document" checked and say "Add question text to Q7 from original document." Download Excel and manually add missing text if needed.

Recode values may be non-sequential (1, 2, 5 instead of 1, 2, 3), missing entirely, or incorrectly assigned during conversion. Smart Fix doesn't change recode values, so use AI Editor: "Fix recode values for Q5 — make them sequential 1, 2, 3, 4, 5" or download Excel and manually edit the Recode Value column.

Yes! Download the Excel file, fix any issues directly (question text, options, recode values, matrix ordering), save, then re-upload to Step 2 for platform conversion. This gives you complete control and is often faster than using AI Editor for multiple complex fixes. Changes made in Excel are permanent.

Pirai AI converts survey structure and questions but doesn't program branching logic or display conditions — you add those in your target platform after import. For example, upload converted survey to Qualtrics, then use Qualtrics' Logic Builder to add "Show Q5 if Q3 = 'Yes'" rules. Logic programming is platform-specific and done post-import.

Pirai AI sometimes uses the first Statement as the Question Text when your original document doesn't have clear matrix question text. This prevents validation errors (matrix questions require Question Text), so the first statement becomes the question and remaining statements are matrix rows. Check your original document — if it lacks explicit question text above the matrix, this is expected behavior.

Yes! Use AI Editor and say "Add question text to matrix Q5: 'Please rate the following aspects'" — this will insert proper Question Text and move the first statement back to the matrix rows. Or download Excel, add text to the Question Text column in Q5's first row, and ensure all statements remain in matrix rows below.

If your original document's matrix lacked a header/title, Pirai AI automatically uses the first Statement as Question Text to create valid Excel structure. This is intentional to avoid validation errors (every question needs Question Text). Use AI Editor to replace it: "Change Q5 question text to 'Rate our services'" if you want different wording.

Compare the converted Excel Question Text with your original document — if the Question Text matches your first matrix statement word-for-word, it was likely auto-generated from the first statement. If it matches a header/title above the matrix in your document, it's from the original. Use "Include original document" in AI Editor to verify and correct if needed.

Google Forms App Script fails when Question IDs contain dots (periods). Pirai AI generates Question IDs based on your document's numbering, which may include dots like "Q1.1" or "Section.2.3". Before running the App Script, you must replace all dots (.) with underscores (_) in the Question ID column. For example: change Q1.1 to Q1_1, change Section.2.3 to Section_2_3. Open your Excel file, find/replace all "." with "_" in the Question ID column, save, then re-generate the Google Forms script.

Google Forms App Script fails when a single question has duplicate answer options. This happens when conversion errors cause the same choice text to appear multiple times in one question (for example, Q5 has "Agree" listed twice as separate options). To fix this: Option 1 (Recommended): Edit your Excel file to remove the duplicate answer options — keep only one instance of each unique choice text per question. Option 2: Comment out the problematic question in the App Script by adding // at the start of the error line, then manually create that question in Google Forms. Option 3: Remove the duplicate question entirely from the script. The only permanent solution is to ensure all answer options within each question are unique in your Excel template before generating the Google Forms script.