BRIEFING · HELP CENTER · AI-POWERED ASSISTANT
Help Center is Threatmatic's own assistant, wired into an org's live data — and which model actually thinks for it is a decision that belongs to the customer, not to us.
Help Center isn't a canned FAQ bot bolted onto the console. It runs with real, live access to an org's own policies, devices, zones, and audit trail — the same data an admin would otherwise have to go dig for by hand, across multiple screens. What model actually reasons over that data, and where that reasoning happens, is the part that's left to the customer.
“We can't send this data to a model we don't control.”
Facet 1
Ask it what it would otherwise take three screens and a support ticket to find out.
you which devices still have TeamViewer allowed? help center 4 devices, all in the Austin office zone. None have been used to connect out in the last 30 days.
Not a search over documentation — a real query against the org's current state, answered in the same sentence it was asked.
Facet 2
Help Center holds real tool access — to policy, device, and π configuration — the same actions an admin could take by hand, gated by the same permissions that already govern the console itself.
The difference between an assistant that tells you what to click, and one that can actually do it, with a permission check in between rather than an extra tab.
Facet 3
Which model actually powers Help Center's reasoning isn't a decision Threatmatic makes for you. Existing vendor relationships, data-residency requirements, cost, capability tradeoffs — whatever the deciding factor is on your side, the choice is yours to make, not something bundled in and locked.
Facet 4
For orgs whose data can't leave a boundary they control — regulated environments, security-sensitive fleets, anyone who's already answered "no" to a shared multi-tenant AI service on principle — a fully private, walled-off deployment option exists. The reasoning stays inside the boundary you set, not a shared service outside it.
An assistant with real access to security-sensitive org data is only as trustworthy as the boundary around where that data actually goes. Threatmatic controls the tooling — what Help Center can see, what it can do, and the same permission model that already governs the console around it. What we deliberately don't control is which model sits behind it, or where that model runs. That decision touches a customer's own compliance posture, vendor relationships, and risk tolerance — not ours to make on their behalf.
Filed · Help Center · console assistant
← Back to The Wire