AI review and assistance

How Ask Alchex answers from your own records, how a document audit grades clauses, and why every AI change waits for your approval.

Alchex has two AI surfaces: a conversational assistant that reads your workspace, and a document audit that grades a document clause by clause. Both work from your own records — documents, controls, risks, audit history — not from general knowledge.

Ask Alchex

Open the assistant with the Ask Alchex bar or ⌘J from anywhere. The panel can Float over the page or Dock to the side, and Esc closes it. Starter prompts cover the usual questions — Compliance gaps, Open findings, Compare audits, Risk overview, Clause coverage, Generate policy.

By default the conversation follows whatever you have open — the input shows Following your page. Focus the conversation pins it instead: pick specific Documents, Risks, Controls or Audits, or leave it on All documents, where retrieval decides what is relevant. Opening the picker lists your most recently updated records straight away, so you can browse and pick without knowing a title; typing filters them, and each tab shows how many matches it holds. You only ever see records your teams can reach. You can also pull a record straight into a message with @. Once you pin something the header reads Focused — not following your page, and Follow page instead returns it to normal.

Answers carry a Grounded in footer listing the records they were built from. If an answer is wrong, use Flag this response as wrong and add a short note; Retry this message re-runs the same turn. Edit & resend from here rewinds — the banner warns that sending rewinds the chat from that point, and everything after that message, including anything the chat had remembered, is dropped.

When the assistant needs input it asks rather than guesses: a question card offers Suggested answers plus Other…, tracks {answered} of {total} answered, and you press Send answers. Facts it learns are shown as Saved to organisation profile with an Undo. Recorded in this chat lists everything the conversation captured, and each line can be removed or forgotten.

You can Dictate a message instead of typing, Stop generating mid-answer, pin an important card with Pin to Decisions, and Share a read-only public link to the conversation.

You never have to wait for a reply to keep going. Sending while the assistant is working adds the message to a Queued strip above the composer — each queued message runs automatically, in order, when the current reply finishes. A queued message is still a draft until it fires: click it to edit it in the composer, use the arrows to reorder, or × to remove it. Pressing Stop generating also pauses the queue — nothing fires until you press Resume — and Clear all empties it. The queue also pauses itself rather than failing through its remaining messages: after a failed turn it holds until you resume, and if the workspace's monthly AI allowance runs out the strip says so plainly — queued messages hold until the allowance is raised or resets. The queue lives in the current chat only; switching conversations discards it.

When a conversation is producing several documents, a Document builds rail appears above the composer listing every build and its state — Drafting…, Ready for review, Created, or Failed — so you don't have to scroll the transcript to check on each one. Review jumps to a build's card; a created document gets its Open document link right on the rail. Builds keep running while the panel is folded, and a build that finishes behind the fold lights the Ask Alchex bar the same way a finished answer does.

Drafts that are ready also count toward the Approve all strip: when two or more proposals are waiting — approval cards, ready document drafts, or a mix — one click approves them all, running each card's own approve exactly as if you clicked them one by one. A proposal the consistency check flagged never joins the sweep; its acknowledgement stays an individual decision on the card.

Attachments

You can hand the assistant a file with the message. Use the paperclip in the composer, drag files onto it, or paste an image straight in. Photos (PNG, JPEG, GIF, WebP) and documents (PDF, Word, plain text, Markdown, CSV) are accepted — a photo of a whiteboard, a signed form, a supplier's report — and the assistant reads them as part of the question rather than as a separate step.

Attach up to four files per message, each photo up to 5 MB and each document up to 20 MB, and 25 MB across the message. Files upload as soon as you pick them, so each one shows as a chip with its name while it uploads; remove one with its ×. The message waits until every upload has finished.

Attachments are stored in your workspace, counted against your storage, and stay on the message — photos appear as thumbnails in the transcript and documents as chips you can open. Only people in your workspace can open them.

A shared conversation is the exception: a public link lists the file names only. The files themselves are never served from a share link.

Nothing changes without your approval

The assistant never edits your workspace on its own. Anything it proposes arrives as a card headed Review & approve this change, with Approve & apply and Reject. Declining says so plainly: Dismissed — nothing was changed.

Cards exist for creating a document from a template (with a read-only Preview document of exactly what will be created), adding a whole document set, logging a Register record, editing a document body, archiving, submitting for approval, publishing, and starting an audit. Document edits show a diff first; if the document moved on underneath, the card offers Re-preview or Regenerate rather than applying a stale edit.

If the assistant describes a change but attaches no proposal, the chat calls it out and offers Produce the proposal — so a described change never gets mistaken for a made one.

Document audits

Open a document from Governance and choose Run audit. The audit reads the document against the clauses in its scope; if none are set, it tells you to add them in Compliance first. Runs finish in seconds and keep going if you navigate away — Audit complete — it finished while you were away.

Each clause gets one verdict: Conforming, Gap, Needs evidence, Not assessed or Excluded. The report records what it was Checked against, the clauses checked, the run date and duration, conformance, and the issues logged. Findings you agree with can be promoted with Log as non-conformity, which files them against the document.

Every report carries the same line: AI readiness assessment — not a certification audit or an auditor's opinion. Treat it as preparation, not as a verdict.

Related