An agentic operating system for how your firm actually works.
Most firms have agents. They don't have an operating system. Aikin maps your firm, decides what stays human, and builds the rest — delivered in tandem with your team.
What an Agentic OS is
An agentic operating system is the layer above your agents — the integrations, data foundation, outcome tracking, governance, and custom software that turn a collection of agents into a system the team can run and trust.
It's two things together.
The agent workflows
The custom software for the missing pieces
Where the custom software runs — your existing stack, EU sovereign cloud, or fully on-chain — is an architecture decision Aikin makes with you, steered by the sensitivity of the data.
How an engagement works
Step 1 — Discovery & Blueprint
2–3 weeksWe shadow your team. We map how the firm actually works — not the org chart, the real workflow. We identify the highest-leverage agents to build, the integrations needed, the data foundation, the governance, and the success metrics that say it's working. The blueprint includes a quoted price and timeline for the build.
Deliverable: An Agentic OS blueprint specific to your firm. Yours to keep, whether or not we go on to build it.
Fixed price: €10–15K.
Step 2 — Build & Implementation
quoted after discoveryThe blueprint becomes a working agentic operating system — agent workflows, integrations into your tools, custom software for the missing pieces, governance, training, handoff. The build is scoped and priced from the blueprint, not before it. Each agent goes live with a human-approval step and graduates to autonomous only as your team's trust builds.
Deliverable: A working Agentic OS.
Price: €40–120K depending on scope.
Step 3 — Optional retainer
ongoingOnce the system is in place, the retainer keeps it tuned as the business, its tools, and the underlying models evolve around it.
€3–8K / month, scaling with the depth of the system.
Built for four verticals
Aikin has engagement plans drawn up for four professional-services verticals. Each one starts from the pain that vertical's buyer actually feels, with a lead agent designed for that workflow.
UX / design agencies
Boutique law firms
Accountancies
Real estate agencies
Governance & defensibility
Every consequential decision, defensible.
An Agentic OS keeps a person accountable for every consequential decision — and keeps a record that holds up when a client, an auditor, or a regulator later asks how a decision was made. Where your work is regulator-watched — and the EU AI Act's high-risk rules reach professional advice, legal interpretation, and case research from August 2026 — that governance is designed into the system from the start, not bolted on after.
What the rules require
The EU AI Act puts obligations on the firms that use AI in high-risk work — and a firm building on top of a model can itself become a provider, with heavier duties still. Either way, the obligations are concrete:
- Maintain logs of AI system decisions and outputs
- Ensure meaningful human oversight of high-risk AI
- Report serious incidents to authorities
- Conduct Fundamental Rights Impact Assessments (for some categories)
- Maintain technical documentation and transparency records
Internal logs vs cryptographic proof
Most companies plan to meet AI Act obligations with internal logging — database entries, audit trails stored on their own servers. The problem: internal logs can be altered. When a regulator or counterparty asks for proof of compliance, you're offering trust, not verification.
Cryptographic receipts on tamper-proof infrastructure can't be altered. They're verifiable by any counterparty — regulators, auditors, customers — without trust assumptions.
A capability, added where it's warranted
Cryptographic provenance, where the workload demands it
Provenance isn't a separate product — it's a capability Aikin folds into an Agentic OS when a workload's sensitivity or regulatory exposure calls for it. Where a decision has to be provable later, the system records it on tamper-proof infrastructure:
Immutable inference logs
Model version attestation
Human oversight records
Compliance dashboards
Cryptographic receipts
Built on ICP's tamper-proof state, with personal data kept off-chain or on a Cloud Engine with a signable operator — the same architectural judgment Aikin brings to sovereign-cloud workloads.
Designed with Okapion
Okapion — a Rotterdam human-centred-design agency — is Aikin's design partner for Agentic OS, co-designing the engagement pattern and lined up as the first external build.
The choice isn't accidental. They had already tried AI — publicly. Their blog post “We had Claude build a platform from Figma — here's what went wrong” (April 2026) is the canonical version of the wall sophisticated firms hit. The Aikin engagement starts where their experiment stopped.
Ready to map your firm?
Discovery calls are 20 minutes. We'll walk through how your firm operates today, where the wall actually is, and whether a discovery & blueprint engagement makes sense as a next step.