Governance for AI-powered operations
FixControl is the governance layer for AI-assisted support and engineering. It runs on top of the tools teams already use, so AI can take on operational work while people keep control of what ships.
Operational work is fragmented, and AI made the gaps obvious
Support, engineering, and deployment each live in their own tools, and people carry context between them by hand. Adding AI without a control layer only speeds the fragmentation up. FixControl closes these gaps with governance rather than another silo.
Fragmented tooling
Support desks, issue trackers, chat, CI, and deployment each hold a piece of the truth. No single place shows the whole operation.
Support disconnected from engineering
A customer problem and its fix live in different systems, so the path from a ticket to a verified code change is manual and lossy.
AI without governance
An agent that can act on production systems without policy, approval, or a record of what it did is a liability.
Poor traceability
When something ships, it is often unclear who decided it, on what evidence, and what changed as a result.
Manual handovers
Context is retyped and re-explained at every boundary between teams. That is slow, and detail gets lost along the way.
Slow incident response
Under pressure, teams stitch together tools and threads by hand — exactly when a clear, governed path matters most.
The rules FixControl is built on
These principles are design constraints. They decide what the product does and what it refuses to do.
Built on top of existing tooling
FixControl connects to the systems you already run — Slack, Teams, Jira, Freshdesk, GitHub, GitLab, Argo — rather than asking teams to move into a new one.
Human approval for high-risk work
By default, a code change, an external reply, or a deployment waits for a recorded human decision before it happens. Any autonomy is explicit, opt-in, and policy-bounded.
Explainable AI
Every proposal comes with the reasoning and evidence behind it, so the person approving understands what they are approving.
Evidence first
Proposals arrive with the facts that justify them. That evidence is frozen with the decision, exactly as the approver saw it.
Governance by default
Policy, verification, and approval are on by default. There is no fast path around the controls.
Auditability
Every meaningful action records who approved it, under which role, on what evidence, and what followed. The result is an audit-ready trail.
Kubernetes and GitOps friendly
Delivery is pull-request and GitOps native. Changes ship through review and your existing pipelines, never a silent push to production.
Tenant isolation
Each customer's data is isolated at the database boundary and enforced fail-closed, so tenants can never see each other's work.
Enterprise security
Signed webhooks, secret-safe logging, scoped admin access, and encrypted credentials come standard.
One pattern kept repeating: lost context, unclear decisions
FixControl came from watching the same operational failures repeat across software projects and organizations: in support, in engineering, in deployment, and in how those workflows fit together.
The pattern was consistent. Capable teams were slowed by fragmented tools and manual handovers, and more recently by AI bolted on in ways that removed control rather than adding it. FixControl is built around one conclusion: make AI useful inside enterprise workflows while keeping humans in control.
FixControl is built by Van Uden Investments B.V. in Hoorn, the Netherlands.
- ObservationFragmented tools, lost context
Very different teams hit the same problems: fragmented workflows, context lost at every handover, and no clear line from a problem to a verified fix.
- The AI shiftAgents could act, ungoverned
As AI entered these workflows, the gap became concrete. Agents could act, but nothing verified, approved, or recorded what they did.
- The goalUseful AI, with humans in control
The aim became a governance layer: let AI propose and prepare the work, but keep policy, verification, and human approval in the path of anything consequential.
- FixControlGovernance became the product
That layer became FixControl. AI drafts the work, policy and verification check it, and a person approves it. Every decision is recorded and auditable.
One governed layer across the whole operation
The direction is a single governed layer across support, engineering, and deployment. AI takes on more of the work; accountable people keep the decisions.
- 01
Connect existing systems
Bring support, engineering, and deployment tools into one governed flow, so work moves between them without manual re-entry.
- 02
Improve operational context
Carry the full context of a problem — evidence, history, and decisions — alongside the work, so nothing is lost at a boundary.
- 03
Govern AI execution
Keep policy, verification, and approval in the path of every AI action that touches a production system.
- 04
Speed up incident response
Give teams a single path to follow under pressure, with the same checks in place, so responding faster does not mean cutting corners.
- 05
Maintain human oversight
Keep a person accountable for every consequential decision, with the evidence they need to make it well.
AI should increase confidence, not reduce control.
Every design decision in FixControl comes back to this. AI in operations is worth adopting when the people responsible can see what it does and stay in control of it.
FixControl is a trade name of Van Uden Investments B.V. The company builds one thing: a governance layer that keeps people in control of AI-assisted operations.