AI agents do the work.
Humans stay accountable.

OctoView Labs builds Topos, an agentic operating system. Agents are the processes. Human signatures are the permission layer. Others record what agents did. Topos refuses what they must not do, and keeps the receipt either way. An append-only ledger records everything, preserved as evidence a regulator will accept. Delivered as an enterprise platform, self-hosted or managed, with the full audit ledger behind a typed API.

Request early access See it in production

Three live views of the Topos substrate: activity map, codebase butterfly, dependency constellation — The live substrate watching its own agents work. The red banner: in one 24-hour window of our own development substrate, the doors **refused 141 claims** from AI agents and admitted the rest. Every refusal is on record with its reason. A system that never refuses anything is not checking. Most systems hide their losses. Ours is architecturally required to confess them.

Named human authorization

No agent act reaches production on its own word. Policy changes carry an operator's signature; the machine signs facts, never permissions.

Independent verification

"Done" is a verdict, not a claim. Separate verifier agents re-derive every result; adversarial tribunals rule on designs before they ship.

Tamper-evident evidence

An append-only ledger is the single source of truth. Append-only means a wrong entry can never be deleted, only superseded. A backup can be restored to hide a mistake; this ledger cannot. That is what an auditor requires, and what accountability demands.

Forged in regulated finance

Our engineering discipline runs EMIR Refit trade reporting live at REGIS-TR since May 2026 for a Swedish financial institution: full trade lifecycle, counterparty delegation, reconciliation, with pipelines on Google Cloud. Topos is the substrate that discipline grew into: it governs the AI agents that build and maintain such systems, and it is built by its own agents, under its own rules, releases included.

Team

Payam Madjidi, Founder. Architect of the EMIR Refit reporting system live at REGIS-TR. Two decades of tier-1 financial infrastructure engineering (TP ICAP, IBM; KTH Royal Institute of Technology). LinkedIn

Johan Önneberg, Co-founder. Leads the commercial side: business model, partnerships, and the conversations that turn infrastructure into engagements. LinkedIn

Stefan Sundkvist, Co-founder. Systems architect; keeper of the hard questions about what a verification chain must prove. LinkedIn

OctoView Labs is a four-person product team based in Stockholm, Sweden. Company page · more on the company page

The question is changing

Everyone will have self-building AI. The question won't be can your agents build?
It will be: can your system testify?

In banking, trade reporting, and law, the deployment decision belongs to the second question. Topos exists so the answer is yes, with evidence to show for it.

We argue this position in public, with receipts, as we build. The laws Topos enforces are published in full on the approach page. The essays cover the law that makes self-modifying machines governable, what happens to engineering work when generation is cheap, and the audit of our own system we published rather than buried. What Topos does for financial institutions and legal work each have their own page, the questions we actually get are answered plainly, and who we are is on the company page.