You are using tokens. Is it driving value?
Token counts and how often people log in are vanity metrics. Real return comes from using AI well — the right skills, in the right hands, Connor puts usage, adoption, cost, risk and the shadow AI across every vendor into one report.
Find out how well your team is using AI.
Through a combination of connections to the Claude and OpenAI APIs, and a scan of every machine in scope, Connor finds the skills and MCP servers your team is using, how many people are using them, and what risks you are exposed to. The report is the output of that work, and it is what you start with.
Who is using it, and how deeply?
The funnel from everybody in scope down to the people through the gateway, and where the drop happens. Usage by product over time, so you can see whether it is growing or whether one launch flattered a quarter.
Who is actually good at this?
The people running the most distinct skills and MCP connections. Not who talks about AI most, but who has built the most that works — ranked, with the counts behind the ranking.
What are we exposed to?
Governed connections against ungoverned ones, the high-risk servers and how many people they reach, and secrets sitting in plain text on laptops. With the shape over time, so you can see whether it is being brought under control.
What is it costing us?
Every seat, every plan, every vendor, and the API usage on top — as one number rather than four invoices. Including the seats nobody is using, which is the fastest saving on the page.
What is Connor giving back?
Tokens removed by Distill before the model ever saw them, at your own configured rate, measured on calls that went through the gateway rather than estimated.
So what do we do?
Every finding turned into something somebody can do this week, tagged with what it improves: publish the skill 132 people rebuilt separately, route the high-risk connections through the gateway, rotate the keys in the open, talk to your three champions.
The report is the baseline, not the deliverable.
A number on its own tells you where you are. What makes it worth having is the second one, three months later, against the same measure — and the work in between that moved it.
- 01
We walk it with you
A session going through what came back, what it means, and which of it is worth doing something about first.
- 02
You act on a handful of things
The recommended actions are already ranked and already specific. Most firms start with the duplication, because it is the cheapest and the most obvious.
- 03
The next report tells you if it worked
Same measures, same sources, three months on. Adoption up or flat, spend up or down, exposure closing or not. That comparison is the thing you take to the board.
If your organisation also has a compliance obligation, the same record answers it — the activity feed, the watch list and the audit trail are already there. That is a different conversation and a different team, and it can wait until you want it. Compliance monitoring.