AtisboAtisbo Docs
Integrations

Paste or upload

Paste text or upload CSV, TXT, and Markdown files to turn research you already have into product signals

Paste or upload

This is the fastest way to bring existing research into Atisbo, and the one worth doing first: interview notes, call summaries, survey exports, or three complaints you remember verbatim. Nothing to authorise, nothing to configure.

Prerequisites

  • An active Atisbo account
  • Either some text to paste, or files in CSV, TXT, or MD format

Two ways in

Paste text. Go to Settings → Sources, open the Paste or Upload card and use the Paste text tab. This is the faster of the two and the one a brand-new workspace needs first — you do not need a file to start.

Upload files. Same card, Upload file tab. Drag files in or choose them from your computer.

Your coding agent can do this for you. If you have a feedback export sitting on disk, ask it to ingest the file — it reads it and sends the contents without you opening Atisbo at all.

Supported File Types

FormatHow Atisbo reads it
CSVOne signal candidate per row
TXTOne text file becomes one uploaded item
MDOne Markdown file becomes one uploaded item

CSV Tips

  • Keep each row focused on a single piece of feedback
  • Export plain text columns when possible
  • Remove purely administrative rows before uploading

If your CSV comes from research tooling or survey exports, Atisbo will still classify and cluster the text after upload.

What Happens After Upload

  1. Atisbo parses each file into signal candidates
  2. Each item is classified as a problem, idea, question, or unclassified
  3. Similar signals are grouped into Claims in Triage
  4. If your workspace does not have an LLM API key yet, the upload is queued and will resume automatically once a key is added

Best Practices

  • Upload raw interview notes instead of only high-level summaries when you have them
  • Prefer smaller, cleaner files over one giant mixed export
  • Split unrelated datasets into separate uploads so Claims stay easier to review
  • Use this for backfills and ad hoc research; connect a live source for anything ongoing, so you are not the mechanism that keeps evidence flowing