Leave with a file, not a transcript

Retrieval was never the hard part. Finding the clause was never what took the afternoon. What took the afternoon was turning what you found into the spreadsheet, the memo, or the audit table that somebody else was waiting on.

Most document AI stops at a good answer in a chat window and leaves that part to you. docAnalyzer treats the file as the point.

Any turn can hand you a file

An answer does not have to be prose. When the work calls for a file, the assistant builds one and it turns up in the conversation as a download:

• Spreadsheets (Excel, several sheets if the work needs them) for extracted fields, comparisons, and tables
• PDFs with real pagination, margins, headers and footers, in the page sizes and orientations you would expect
• Web pages that carry their own images and styling inside the single file, so nothing breaks when you email it to someone
• Charts and diagrams drawn from your data, including flowcharts and process diagrams
• Mathematical notation, typeset properly rather than described
• ZIP bundles collecting several of the above, filenames preserved

Any file it has built can be converted afterwards, so a report becomes a Word document and a spreadsheet becomes a PDF without anyone rebuilding it.

One thing to expect from the interaction: the assistant decides when a file is the right answer, based on what you asked. There is no menu where you pick "make me a PDF". You describe the work, and if the result belongs in a spreadsheet it arrives as one.

Work accumulates instead of resetting

This is the part that changes how a long session feels.

Every file produced in a chat stays available for the rest of that chat, and the assistant can still reach it. Extract data into a spreadsheet early on, keep working, and eleven questions later you can ask for that spreadsheet as a PDF, or bundled with a chart you made since, or rebuilt with two columns dropped. The assistant passes the file it already built into the next step. It does not regenerate the content and hope it comes out the same.

The practical difference: a session builds a body of work instead of producing a series of disconnected answers you then reconcile by hand. Ask for a revision and you get a revision of the thing, not a fresh attempt at it.

Run it across the set, not one document at a time

A Workflow is the same engine run in bulk. Instead of a conversation, you point one at a set of documents and it performs the same task on every one of them, handing back a single organised result. Six of them ship today:

• Data Extractor pulls the same fields out of every document into a spreadsheet, one row per document, one column per field. Tell it which fields you want, as a list or in plain language.
• Summarizer summarizes each document at the length, tone, and format you specify.
• Blueprint audits each document against a reference document or a checklist and reports gaps, matches, and mismatches.
• Individual asks the same question of every document and returns per-document answers.
• SEO Metadata generates page titles, descriptions, keywords, and the preview tags that decide how a link looks when it is shared.
• Humanizer rewrites AI-generated prose through targeted edits while holding the facts steady.

Each document is handled separately and results are saved as they land, so a run survives a browser refresh. Start a run across a few hundred contracts, close the tab, come back and inspect, collect, or cancel it.

Reshape a finished run without running it again

A workflow returns its result in one fixed shape, which is rarely the shape you need to send to someone.

Open a completed run in a chat and the result loads into a panel with the chat box underneath it. From there you can ask for the rows built into a spreadsheet, a chart drawn from them, columns dropped or reordered, the whole thing turned into a PDF, or several files bundled together. The chat reads the result directly, so there is no copy-paste, no re-upload, and no second run against your credits.

If one of the run's per-document files is already the shape you want, download it from the panel and skip the chat.

What you should know before you rely on it

Files live with the chat that made them. They stay available for as long as that chat does, up to 12 of them, and they are not filed away into your workspace alongside your documents. Download what you intend to keep before you close it.

Saving an answer as a Note keeps the text, not the file. A Note is a writable document that lives in your workspace, and "Save as note" puts the answer's wording into one. The spreadsheet attached to that answer is a separate thing, and downloading it is a separate action.

Composed files have size ceilings and spreadsheets carry the usual Excel constraints, including the 31-character sheet-name limit. Large builds can hit those edges.

You are directing, not operating. Because the assistant chooses when to build a file, phrasing matters. "Put the deviations in a spreadsheet with the clause and the contract name" gets you a spreadsheet. "Tell me about the deviations" gets you prose.

Where this leaves you

The deliverable is the unit of work. A session that ends with a spreadsheet your colleague can open, an audit table with citations that resolve, or a PDF you can attach to an email is a session that finished the job rather than describing it.

The grounding behind those files is a separate question, and a fair one to ask. How docAnalyzer grounds its answers covers where a citation actually points and what the retrieval is doing.

Start your first analysis on the Community plan. No credit card.