Ask the demo archive

A chatbot can only see what you paste into it. This demo starts where that ends: download a 213-document archive, feed it back in, and ask the questions that need all of it — trends, cross-company comparisons, what’s missing. Everything runs in your browser and against our pipeline; nothing you ask or upload is stored by us. No signup, no cookies.

1
Download the archive

213 files, nine companies, zipped. Try pasting them into a chat window first — that’s the point.

Download demo-corpus.zip ↓
2
Feed it back in

Drop the zip below. Ingestion runs in your browser — you’ll watch every document go in.

3
Ask across everything

Gap detection, trends, comparisons — answers a context window can’t give, each showing what was searched and read.

213documents
9companies
2019–2026date span
Where your documents actually live
DocuStrataOCR · date · entity · index
Sources on the left · the filed archive on the right · click any page to read it

Your file is read in your browser and never touches our storage — its text exists only inside the question you ask. Processed by Anthropic under the same commercial terms as the product: never used for training, deleted from its systems within 30 days.

Five questions an hour · the demo rests when its daily budget is spent · answered by the same retrieval-then-read pipeline as the product, capped to its smallest model.

Now imagine it on 25,000 of your own documents.

Start free — no card required