Mission

Intelligence that can be audited by the people it affects.

Most systems ask you to trust a black box. We build the opposite: a published corpus, sourced answers, append-only records, and reasoning any outsider can replay. Not because trust is bad — because it should be earned in public.

What we are for
01 · Audit

Checkable by outsiders

A published corpus, sources under every answer, and a record that only ever grows. Nothing here needs to be taken on faith.

02 · Access

For people without a staff

Built for the people who don't have lawyers or analysts standing by — the same intelligence, without the retainer.

03 · Restraint

Honest about the edges

Live, in beta, or not built yet — each surface says which. A roadmap dressed as a product is a lie you'll find out about later.

Commitments, and how you check them

The reasoning is published

The corpus Elle stands on is publicly indexed and citable, including the parts that are unfinished or wrong.

PhilPapers index
Answers carry their sources

Not a confidence score — the actual documents, listed under the answer, openable.

Sources on every answer
Records are append-only

Corrections stack on top of mistakes. A wrong entry stays wrong, in public, with the correction above it.

Optimus journal · run replay
She cannot ship her own code

Elle writes code and CI judges it, but the merge is a human click. No merge tool exists for her to call.

Forge → CI → human click
Your records are not a dataset

Not sold, not trained on. Export everything or delete the account and everything in it, any time.

Export and delete, any time

Sign in

One free account for Elle, your Optimus journal, and the analysis engine.