How to use Pitch Deck Analyzer

Reference — comprehensive feature documentation.

What is Pitch Deck Analyzer?

Pitch Deck Analyzer is an AI-powered Labs feature that analyzes the pitch decks and investment documents you upload to Affinity. When you click Analyze, it provides structured summaries of key investment information, delivers recent news about the company, and suggests CRM field updates to keep your records current—all designed to help you make faster, more informed investment decisions without manually reviewing every slide. The workflow: Upload PDF → Click Analyze → View summary + news → (Optional) Suggest field updates → Apply updates What it eliminates: Hours of manual deck review and research. The AI reads the entire document, extracts key insights, searches for recent news, and identifies CRM fields that should be updated based on document content. Who it’s for: VCs, PE investors, and deal teams who review pitch decks, CIMs, financial reports, and other investment documents regularly. Where to access: Files tab on any company, person, or opportunity profile (available on web and on the Affinity iOS app for iPhone and iPad) Feature status: Active Labs experiment (launched December 2025)

Step 1: Upload a Document

Navigate to the Files tab on a company, person, or opportunity profile
Upload a PDF file:
- Click Upload or drag and drop
- Supported: PDFs up to 23 MB
- Works with pitch decks, CIMs, financial reports, strategic plans, research papers
File appears in the Files tab

Step 2: Analyze the Document

Find the uploaded file in the Files tab
Click the Analyze button on the file card
Analysis modal opens and begins processing:
- Initial analysis: 10-30 seconds
- Recent news search: Additional 5-10 seconds
- Total: ~30-60 seconds for most documents
You can close the modal and navigate away—analysis continues in background

Progress indicators:

“Analyzing document…”
“Searching for recent news…”
“Analysis complete”

Step 3: Review the Analysis

The analysis modal shows: Document Summary:

Key investment insights extracted from the deck
Structured sections (business model, market, traction, team, etc.)
AI adapts sections based on actual document content
- Pitch deck: Shows company overview, product, market, traction
- Financial report: Focuses on performance metrics, financial health
- CIM: Highlights investment opportunity, terms, key risks
Only relevant sections appear (no empty templates)

Recent News (Additional Research):

Links to recent news articles about the company
Article source and publication date
Snippet from the article
Explanation of why the article is relevant
Refresh button: Regenerate news search for latest articles

Field Suggestions (Optional):

Click “Suggest Field Updates” button
AI extracts structured data from document (funding, revenue, stage, location)
See before/after comparison for each field
Each suggestion shows source slide/page number
Select which to apply, click Apply to batch update

Step 4: Apply Field Updates (Optional)

After reviewing the analysis, click “Suggest Field Updates”
Analysis takes additional 5-15 seconds
Review suggested updates:
- Current value vs. suggested value
- Source page/slide where data was found
- Select/unselect individual suggestions
Click Apply to update selected fields
Fields update immediately on the profile

You control what gets updated—nothing happens automatically.

What Types of Documents Work?

Supported file types:

PDF files only
Up to 23 MB file size

Document types that work well:

Pitch decks (primary use case)
CIMs (Confidential Information Memorandums)
Financial reports (quarterly/annual performance)
Strategic plans (company roadmaps, initiatives)
Research papers (market analysis, whitepapers)
Legal documents (term sheets, contracts)
Product documentation (technical specs, architecture)

The AI adapts its analysis approach based on document type. A pitch deck gets different treatment than a financial report—sections and insights adjust to what’s actually in the document. Note: While primarily validated against pitch decks, the analyzer provides valuable summaries for all investment-related documents. Share feedback on non-pitch-deck use cases.

Understanding Field Suggestions

Which Fields Can Receive Suggestions?

Supported field types: Text, multiline text, number, date, location (same as Field Updates from Notes) Field eligibility:

Company and Person profiles: Suggestions are generated for global fields (including Affinity Data enriched fields and custom global fields). In addition to being a supported type, the field must be configured to display on the profile by an admin via Settings > Profile Field Customization — global fields that aren’t selected in Profile Settings won’t receive suggestions.
Opportunity profiles: Suggestions are generated for list-specific fields. The Profile Settings requirement does not apply.
In all cases: Fields must be manually editable. Identity fields (name, primary domain, email address) are excluded.

Common fields that receive suggestions:

Funding Stage (“Series B” from deck)
Revenue (“$5M ARR” from financials slide)
Team Size (“25 employees” from team slide)
Location (“San Francisco” from contact slide)
Founded Date (“2019” from company overview)
Investors (“Sequoia, a16z” from funding slide)

How Suggestions Work

Source attribution: Each suggested field update shows which slide or page the data came from. Example:

Field: Revenue
Current: Empty
Suggested: $5M ARR
Source: Slide 8 (Traction)

Before/after comparison: Clear visual diff showing:

What’s in the CRM now (current value)
What the document indicates (suggested value)
Which slide/page the AI found it on

You decide what to apply:

Select all and apply with one click
Or uncheck individual suggestions
No inline editing (if wrong, uncheck and update manually)

Permissions

Your existing permissions are always respected—you can only analyze files you have permission to view.

Limitations

PDF-only support: Currently works only with PDF files. PowerPoint, Word, Google Slides, and other formats are not supported. Convert to PDF first. File size limit: 23 MB maximum Larger files will fail to upload. Compress the PDF if needed. Recent news depends on public coverage: Companies with limited press coverage may show fewer or older articles. The AI prioritizes recency but may surface older news if nothing recent exists. Field suggestions require separate enablement: Field suggestions are a separate Labs feature. Both “Pitch Deck Analyzer” and “Field Suggestions for Files” must be enabled in Labs Settings. Analysis is per-document: You must analyze documents individually. There’s no bulk analysis across all uploaded files.

Troubleshooting

”File too large” error

Cause: File exceeds 23 MB limit Solution: Compress the PDF using online tools or Adobe Acrobat to reduce file size below 23 MB.

Can’t see the Analyze button

Check:

Pitch Deck Analyzer is enabled in Settings > Labs
File is a PDF (not PowerPoint, Word, or other format)
You have permission to view the file

Solution: Enable the feature in Labs Settings. Convert non-PDF files to PDF format.

Additional Research showing old articles

Cause: Limited recent press coverage for the company Why this happens: The AI prioritizes recent news, but if the company hasn’t been in the news recently, it surfaces older articles rather than showing nothing. Solution: This is expected behavior. Click Refresh to re-run the news search in case new articles have been published since analysis.

Analysis seems outdated

Cause: Previously analyzed document showing cached results Solution: Click Refresh in the Additional Research section to regenerate the news search with the latest articles.

No field suggestions appearing

Check:

“Field Suggestions for Files” is enabled in Labs Settings (separate from Pitch Deck Analyzer)
Document contains relevant structured data (funding, revenue, team size, etc.)
Fields aren’t already up to date
For Company/Person: Fields are enabled in Profile Field Customization Settings (admin)
Fields are supported types (text, multiline text, number, date, location)

Solution: Enable both Labs features. Confirm document has extractable data and fields meet eligibility requirements.

Best Practices

Upload high-quality PDFs: Text-based PDFs work better than scanned images. If you have a scanned deck, use OCR (optical character recognition) to convert it to searchable text first. Analyze complete decks: More comprehensive documents yield richer analyses. Incomplete decks or excerpt PDFs may produce limited insights. Save important insights: Export or copy key sections to notes for permanent records. Analysis results aren’t saved long-term—they’re regenerated each time you open the modal. Use field suggestions selectively: Not every deck contains reliable field data. Review suggestions carefully, especially for financial metrics where precision matters. Provide feedback on document types: If you analyze CIMs, financial reports, or other non-pitch-deck documents, share feedback on how well the analysis worked. This helps us improve the AI for diverse document types.

What is Pitch Deck Analyzer?