A few dozen automations are easy to watch. Govern a thousand.
If it takes a team of people to watch the automations, nobody saved any time. Caddi keeps one view of every loop running, being built and discovered across the firm, and replays any run decision by decision with the permission it used.
Replayable runs·SOC 2 Type II·From the Team planRunning, being built, discovered. One list.
The question is not whether one loop worked. It is what is running in my firm right now, who owns it, what is about to join it, and what should.
- RunningEvery loop that runs, with its owner
Name, owner, version, and what it did today. A loop with no owner is the thing this view exists to prevent.
- Being builtEvery loop in Studio, and who is building it
What is coming, before it arrives. The pipeline, without asking anyone for a status update.
- DiscoveredEvery finding, by team
What Discover found across the firm and which team it belongs to, so the work worth automating is a list you can assign.
Every run on the record. No black box.
Deterministic code does the work, so the same input takes the same path every time and the path can be written down.
- Every run, replayable
Open one and read what it decided, in what order, and what it touched to do it.
- Every decision, with its permission
Each line names the system and the scope it used. The step where it decided not to write is on the record too.
- Every stop, immediate
Pause a loop or pull its connection to a system. It stops mid-queue, not at the end of the batch.
The record is yours to pull.
A program running several AI tools needs one report across all of them, and the tool that hides its logs is the one that cannot be in it.
- In the productRun records with counts by system
Records read, documents filed, messages sent, cases held. Legible on screen, and the same data your reporting would want.
- EnterpriseAction-level audit logging
Every action, per system, per run, with the permission it used. Built for the audit, and for the security review before it.
- Your reportingThe record is yours to pull
If your program reporting lives in a warehouse or a dashboard of your own, talk to us about the export.
What a program lead asks
Can a loop decide something on its own?
No. Deterministic code runs the work, AI is used inside specific steps that need judgment, and changing a process takes a human approval. The plan is the boundary, and the plan is written down.
Can we prove what it did, and how fast can we stop it?
Every run replays decision by decision, and every decision names the system and the permission it used. That record is what an auditor reads and what the security review asks for first. To stop a loop, pause it or pull its connection to a system. Either takes effect mid-queue.
Does the firm-wide view show automations that run outside Caddi?
Not today. It covers the loops Caddi runs, the loops being built in Studio, and the processes Discover has found. That last shelf is wider than Caddi, so it will surface work you already automate elsewhere; when it does, treat it as a check that the method is finding the real work. A view across every automation platform a firm runs is something we are thinking about with customers, and it is not on this page until it exists.
How do teams, access and plans work?
Loops and findings roll up by team, and the firm-wide view is on the Team plan. Action-level audit logging, SAML SSO, centralized MFA, SCIM provisioning and centralized integration governance are on Enterprise. Each connected system carries one scope, under the access the person already had. See pricing.
Is our data used to train models?
No. The contents of your connected systems are never used to train models. OpenAI and Anthropic calls run under zero-retention terms, and accounts under a Data Processing Addendum exclude their content from model development entirely.