What Louvain is

Answers your firm can stand behind — cited to the source and limited to what each person is cleared to see — in the app, through the API, and in the AI assistants your firm connects over MCP.

The answer to a client's question is usually somewhere in the firm already: a mail thread from March, a working paper nobody linked, a note in a Slack channel.

Louvain connects those sources and answers questions from them, with the sources attached and every figure, date and name checked against them. Partners and client teams ask in the app; the AI assistants they already use — Claude, Claude Code, Cursor or your own code — ask through the API and MCP and get the same checked answers.

Every answer, whether a person or an AI assistant asked it, is built only from what that person is already cleared to see.

The part that is different

Most tools that put AI over a firm's documents build one index and hope the prompt keeps the secrets. Louvain resolves permissions before ranking: a question is answered against the subset of the corpus the person asking can already see. Two people can ask the same question, in the same deployment, on the same day, and correctly get different answers.

A correct refusal and a missing record look identical from outside. That is deliberate. Anything else leaks the existence of what it is hiding.

Start here

Understand the model

Worth twenty minutes before you connect anything real — the permission and provenance rules are what make the answers trustworthy, and they have consequences you should expect rather than discover.

Connect your sources

Use it

Reference