OduPilot / AI inside Odoo
Work with an AI agent without leaving Odoo.
Bring assistants into the screens where your team already works. OduPilot gives agents their own conversations and workspace, connects them to Odoo through OduMCP and lets people review tool permission requests in the conversation.
What it does
OduPilot in your workflow.
An AI agent platform inside Odoo: conversations in Discuss, questions about records and developer workspaces.
- Conversations in Discuss
- Start an agent chat with its own history and workspace, share files and invite internal colleagues into the conversation.
- Ask about a record
- Ask AI from a saved record's chatter. The question and answer remain as an internal note, subject to the user's normal record access.
- Visible tool decisions
- Review permission cards in the chat, allow or reject requests, and follow live response and tool activity with decisions recorded in the audit history.
- A workspace for development
- Administrators can start an AI Developer request with the current screen context. Configured infrastructure provides an isolated Git worktree for code changes.
How it works
From setup to everyday use.
Choose an agent
Open AI Agent Chat in Discuss and select an agent configured by your administrator, or ask a question from a record.
Work in context
Share the task and relevant files. The agent uses its configured model and the Odoo tools permitted by your MCP profile.
Keep the conversation
Review tool requests and answers in Odoo. Closing the session revokes its AI access while preserving the Discuss history.
Running OduPilot
Requires OduMCP, the OduPilot AI runtime, configured provider credentials and user profiles. Developer workspaces need the corresponding external infrastructure. OduPilot is distributed under OPL-1; see its repository for license terms.
Put OduPilot to work.
Discuss your Odoo setup and how this solution fits your team. Explore the full development and operations platform on the Oduflow website.