Every AI-assisted action needs a decision ledger
FixControl is the governance layer for AI-assisted support and engineering — not a generic agent. Each meaningful action records who approved, which role, which channel, what evidence was shown, what was approved and what followed.
From proposal to a recorded decision
The AI proposes; a person decides; the decision and its consequence are written down.
- 01ProposalAI proposes an action with evidence
Whether it is a reply, a patch or a deployment promotion, the action arrives as a proposal — with the evidence that justifies it laid out for a human to read.
- 02Decision surfaceA person decides on a governed surface
Approvers act from where they work — Slack and Teams approval cards, or in-app — so the decision is captured at the moment and place it is made.
- 03Ledger entryThe decision is frozen into the ledger
On a verdict, FixControl records who approved, their role, the channel, the timestamp, the evidence shown and the exact text, diff or action approved.
- 04ConsequenceThe outcome is recorded too
What actually happened after approval is captured alongside the decision, so the ledger reflects consequence, not just intent.
The full shape of a governed decision
Each entry answers who, under what role, from where, on what evidence, of what — and to what effect.
Who approved, and which role
Every entry names the actor and the role they acted under, so an approval is attributable — not an anonymous machine action.
Which channel the decision came from
The ledger records whether a verdict arrived via Slack, Teams, the in-app surface or another channel — the decision's provenance is part of the record.
What evidence was shown
The evidence presented at the moment of decision is frozen with the verdict, so a later reviewer sees what the approver actually saw.
What was approved — text, diff or action
The exact final reply text, code diff or action that was approved is captured, not a vague summary of it.
What consequence followed
The outcome of the approved action is recorded alongside it, closing the loop between a decision and what it caused.
Timeline, separate from raw logs
An operational timeline tells the human story of a mission, kept distinct from raw agent traces — readable for governance, not just for debugging.
Governed AI, not autonomous agents
Common questions
How is this different from a generic AI agent?+
What exactly is recorded for each decision?+
Where do people approve?+
Why separate a timeline from agent logs?+
Is the ledger built for audit and export?+
Does any meaningful action skip approval?+
See governed AI in action
Walk through approval cards, the decision ledger and the operational timeline on a live demo.