Skip to main content

​
1. Create a Project

After signing in, you’ll be prompted to create your first project. Each project is a self-contained workspace with its own file tree, chat sessions, and review history.
New project IDE view
1

Name your project

Give your project a descriptive name (e.g., “BLA Submission Review” or “Q1 Patent Analysis”).
2

Set visibility

Choose Organization to let all org members access it, or Private to restrict access to invited members only.

​
2. Upload Documents

Drag and drop files into the workspace explorer, or use the upload button. Supported formats:
FormatParsingEditing
PDFAutomatic OCR via MistralView only
DOCXAutomatic parsing via ReductoOnlyOffice inline editing
XLSXAutomatic parsingOnlyOffice inline editing
MarkdownNativeInline text editing
PlaintextNativeInline text editing
Documents are automatically parsed in the background to power AI features like search and chat.

​
3. Chat with Your Documents

Open the chat panel and start asking questions about your uploaded documents. You can:
  • Ask questions — “Summarize the key findings in this protocol”
  • Reference files — Use @mentions to attach specific documents to your message
  • Generate content — “Draft a response to this RFI based on our SOPs”
  • Run code — The AI agent can execute bash and python for data analysis

​
4. Review Documents

Set up structured review workflows:
1

Create a Review Program

Define the review prompt and attach guideline files that the AI should follow.
2

Run a Review

Select target documents and kick off the review. The AI analyzes each document and produces action items.
3

Resolve Action Items

Review findings by severity, resolve items, and assign a final disposition code.

​
5. Commit Your Changes

After making edits (manually or via AI), review the diff in the workspace panel and commit your changes. Every commit is versioned — you can always go back to a previous version.