BTW Studio
← All work

Answers that show their source.

Citewise answers questions from company documents. Every sentence carries a citation, and every citation is checked against the document before anyone sees it.

Open the demo ↗Read the code

In plain words

You ask a question about your company documents and get a short answer. Next to each sentence is the exact line it came from. When the documents do not hold the answer, it says so instead of guessing.

How it works

  1. Ask in your own words

    The question goes in the way a person would type it into a help chat.

  2. Get an answer with its sources

    Each sentence links to the passage it rests on. One click highlights the quoted line in the document.

  3. Or an honest "I do not know"

    A question the documents do not cover is refused. Nothing is filled in from general knowledge.

The problem

A document assistant that sounds right and points at a passage that does not say what the answer claims. Prompting harder does not fix it.

What we built

The answer is drafted as separate claims, each with a source and a word-for-word quote. Plain code then checks that the quote really is in that source. A claim that fails is dropped; if nothing is left, the assistant says it does not know.

How we know it works

Test questions22
Right passage in the top three94.4%
Answered and cited the right passage88.9%
Refused questions the documents do not cover100%
Citations that passed the check100%

Two questions are missed and stay listed on the demo page: keyword search does not connect "lose my laptop" with "lost device".

Built with
TypeScript, Express, own BM25 search
Model
Any OpenAI-compatible endpoint; runs without one
Checks
9 unit tests, eval gate
Good fit for
Help centers, internal handbooks, policy search

Under the hood

  1. DocumentsMarkdown, one chunk per heading
  2. SearchBM25 picks the closest passages
  3. DraftClaims, each with a source and a quote
  4. CheckThe quote must be in the source, word for word
  5. AnswerOnly claims that passed
    or
    RefusalNothing passed

The model path is covered by tests against a stubbed endpoint, including a fabricated quote that must be dropped. It has not been run against a live provider yet.

Decisions and why

The check is code, not a prompt

String matching does not depend on the model behaving. A made-up quote is dropped every time.

Keyword search before embeddings

BM25 needs no outside service, so the demo costs nothing to run. The index sits behind one class and can be swapped for embeddings or a hybrid.

The model is replaceable

It sits behind one interface. Without a key the answer is assembled from the best-matching sentences, so the whole pipeline runs with no model at all.

The misses stay visible

Two test questions fail and are shown on the demo page. They mark where keyword search stops.