Bring workplace AI use into the open.
A shared register of which AI tools a small team may use, what data is allowed in each, and who signed off.
Free plan up to 10 tools; Starter $9 and Team $19 per workspace, monthly.
The problem
Employees adopt AI tools faster than anyone can track, and bans just push the usage out of sight. When a client questionnaire or insurance form arrives, there is nothing to point to.
The approach
ToolRoster keeps one visible register of approved, restricted, and blocked tools with the data rules for each, routes new requests through a recorded approval, and exports the evidence on demand.
What's inside
Tool registry with five statuses and full history
Per-tool data rules, from public to regulated
Deterministic risk scoring from recorded vendor facts
Request-and-approval workflow with recorded decisions
Versioned AI usage policy with acknowledgments
Readiness score and one-file evidence pack
Public approved-tools page, no login required
Append-only audit log with CSV exports
03From the lab
The decisions, trade-offs, and engineering notes behind ToolRoster — written as we build.
We spent an afternoon researching the risk of twenty AI tools and found five questions with no public answer — because they aren't facts about the tool, they're facts about your own company that nobody wrote down.
The template deliberately names no AI tool, because the tool list changes weekly and the policy shouldn't. Here's the full text, the reasoning behind each section, and what we left out on purpose.
We sat down to write the "August deadline" post our own marketing plan called for, checked the current text first, and found the deadline had been amended out from under us. Here's what genuinely lands, what moved, and why the client questionnaire is the harder deadline anyway.
More from Eastbase
