Pilea accepts feedback in many formats — from spreadsheets and surveys to interview transcripts, PDFs, and images. Upload any document containing customer signals, and Pilea automatically analyzes it and creates backlog items from the patterns. You can categorize each upload by content type so Pilea analyzes it appropriately.
What file formats are supported
Pilea accepts:
Spreadsheets: CSV, TSV, XLSX, XLS (with header rows)
Documents: PDF, Word, Google Docs exports, text files
Images: Screenshots, mockups, bug photos (visual evidence of issues)
Transcripts: Interview transcripts, meeting notes, call recordings transcribed
File size: Up to 50MB each. Upload up to 20 files at once.
Pilea auto-detects content in your files:
Text columns (feedback, questions, responses)
Dates (when responses were submitted)
Email addresses (to link responses to existing customers)
Names (associated with respondents or speakers)
Images/attachments (visual evidence)
ℹ️ Note: There's no required column name or document structure. Pilea reads the content and sample values to detect what each column or section contains. If you have a simple two-column CSV with "Customer" and "Feedback", it will work.
Content types — help Pilea analyze correctly
When you upload, Pilea asks you to categorize the file so it analyzes appropriately:
Content Type | Examples | Analysis Intent |
Survey | NPS/CSAT exports, app reviews, form responses | Analyze (extract customer feedback) |
Interview | Customer interview transcripts, usability test notes | Analyze (primary research) |
Sales / CS Call | Demo transcripts, renewal conversations | Analyze (direct customer feedback) |
Customer Meeting Notes | Notes from customer meetings, partner reviews | Analyze (customer feedback signals) |
Internal Meeting Notes | Sprint planning, retros, product reviews | Knowledge (reference material without backlog extraction) |
Feedback List | Bulk feedback exports, bullet-point lists | Analyze (individual items) |
Document / Report | Research reports, strategy docs, insight decks | Knowledge (pre-synthesized context) |
|
|
|
Analyze intent tells Pilea to extract individual customer signals and create or update backlog items. Knowledge intent treats the file as reference material — Pilea stores it and you can search across it, but it won't automatically generate backlog items.
How to upload your files
Go to File Uploads in the left sidebar.
Click Upload Files.
Drag and drop your file(s) onto the upload area, or click to browse.
Select the Content Type that matches your file (Interview, Survey, Feedback List, etc.).
Choose the Analysis Intent: Analyze (to extract backlog items) or Knowledge (to store as reference).
Click Upload Files.
SCREENSHOT: The Upload Files upload area.
💡 Tip: If you have multiple files to import, you can add up to 20 at once in a single upload.
What happens after you upload
Pilea processes the file in the background. You'll see the progress update to Analyzing responses as each row or section is processed. Once complete, the file appears in your File Uploads list.
Click the file to open its detail page. You'll see three tabs:
Summary — an AI-generated overview of the themes and patterns across all content
Responses — a list of every imported entry (rows from a spreadsheet, transcript sections, image descriptions, etc.)
Backlog Items — the backlog items created or updated based on the feedback (only shown if Analysis Intent was "Analyze")
[SCREENSHOT: The file detail page showing the Summary, Responses, and Backlog Items tabs]
Each entry becomes a mention in Pilea, processed the same way as manually added feedback or feedback from an integration. Entries are linked to backlog items, and if an email column or contact name was present, they're matched to customer records automatically.
Viewing the imported entries in Mentions
The imported entries also appear in the Mentions list. You can filter by source to see only your imported entries. Each mention shows its processing status — entries move from Received to Processed once analysis completes.
ℹ️ Note: If your file contained an email column or contact name and a customer with that email already exists in Pilea, the entry is linked to that customer profile automatically. If no match is found, the entry is still analyzed, but it will not be tied to a customer record.