About
Making artificial intelligence accessible and operational
EasyRag grew out of a simple observation: organizations hold a wealth of knowledge in their documents, yet struggle to make it accessible. Language models make this possible, provided they are connected to that content, governed and administered.
What we are building
EasyRag is a software platform for creating, publishing and monitoring AI assistants that answer from an organization’s documents and cite their sources. It covers the entire lifecycle: knowledge import, assistant and model configuration, publishing on the web, by email, in Slack or via API, monitoring and administration.
Its governance features were designed to meet the requirements of organizations, particularly in the public sector: single sign-on including FranceConnect+, data retention, per-organization white labeling and content validation.
A community platform
EasyRag is not a company’s proprietary software: it is a community platform, published byPatrick Beaucamp and distributed as Docker images, which each organization installs and runs in its own infrastructure.
The community is built around the wiki documentation (in French) (opens in a new tab), the how-to guides and feedback from its users: discover the community.
Supporting EasyRag
The EasyRag platform and community are supported by their sponsors, who fund and promote the project.
LaTribu-IT.fr(opens in a new tab)
Main sponsor: the company that funds and promotes the EasyRag community.
Patrick Beaucamp
Founder and publisher of the EasyRag platform and website.
Our principles
Verifiable answers
To be useful in an organization, AI must be able to show where its statements come from. Citations are at the heart of the platform.
Control, not lock-in
Your choice of model, installation in your own infrastructure, data in your own databases, migration between instances.
AI that serves teams
Tools designed for business teams, with safeguards for IT departments and data protection officers.
Honesty about capabilities
We state what works, what requires configuration and what is still to be done. This website is a case in point.
Want to learn more?
A platform walkthrough, a discussion about your context, a partnership: get in touch.