Transaction services teams receive data rooms and shared folders with hundreds of unsorted PDFs, emails, and spreadsheets. Chainlocker reads every file, judges it against your diligence checklist, and returns an Excel index with a tag, written reasoning, and a confidence score for each document. A person confirms every call. Documents never leave your office.
The categories a document can land in and the instructions for deciding: material contracts, financial statements, tax filings, side letters, leases, whatever your engagement letter calls for.
The data room export or the client's shared drive. PDF, Word, Excel, PowerPoint, email exports, and scans, listed and counted before anything runs.
A tag from your list, written reasoning, and a confidence score. Anything below the floor you set goes to review rather than being filed.
Accept by default, reject with a correction. Working a large folder this way is quick, and every verdict is recorded with who decided and when.
An Excel file, one row per document, with a clickable link to the file, plus a self-contained package that opens anywhere with no server and no network.
The criteria list is the standard. Run it on the next data room and the triage is done the same way, whoever staffs the engagement.
Chainlocker runs on a dedicated appliance installed at the firm and is reached from a browser on your LAN or VPN. Documents never leave it. AI inference runs on hardware we own over an encrypted private link, never a third-party AI API, so there is no usage meter and no provider terms of service between you and your client's files.
The deployment model in detail →Chainlocker is in development, targeting late 2026. Before it ships we want transaction services teams running their own checklists through it and telling us where it falls short.