Big firms build ServiceNow. We make sure it's built right — and stays under your control when AI starts building it for you. Kritavi is a senior, Reykjavík-based consultancy for organisations that run on the platform: advisory for the instance you have, and fixed-scope implementation for the one you need.
Whether you already run ServiceNow or are about to, the work is the same discipline applied at different points: structure the platform properly, govern how it changes, and measure what it returns.
For complex and regulated organisations where ServiceNow is a control system, not a ticketing tool. We take the platform owner's side of the table.
Full implementation from data model to go-live, on a fixed scope and a fixed date. Governance and the data model are designed in from the first workshop — not bolted on at the end.
One standard, whichever way in. When another partner builds, we hold them to it. When we build, the same standard holds us — and we never QA our own work and call it independent.
Before we recommend anything new, we run a fixed-scope review of the platform you already have — what your entitlements can already do, what it actually costs to change, whether it survives the next upgrade, and where governance is missing. You leave with the Standard — a written baseline you own — that you can hold your team, your partners, and the AI now changing your platform to. Three weeks. Fixed scope. Fixed price.
A structured read of what you're entitled to versus what's switched on — capability by capability — and the shortest path to using what you already pay for.
Incident, request, case and change processes rebuilt around how your organisation actually works, with the exceptions designed in rather than discovered in production.
Technical-debt and customisation review, CSDM conformance, and a standing upgrade posture — so releases are routine events, not projects.
Written guardrails, per process and per role: who may change what, how configuration is reviewed, what gets measured monthly — and how change made by AI agents is approved before it reaches production. When the platform starts configuring itself, this is the control that keeps it yours.
One document — the Standard. Per process and per role: what "good" looks like on your platform, who may change it, how it's measured, and how change, human or AI, gets approved. The baseline your team builds to, the yardstick you hold your implementation partners to, the control record you can show a regulator. Owned by you — not locked inside a consultant's methodology.
Every hour delivered by a certified senior architect. Fixed scope, fixed price, fixed date.
Other partners build, then offer governance as a service line. We run the practice that holds implementation partners to a standard — so we build to that same standard, and hand it over ready to be audited against it. Fixed scope, fixed price, senior-delivered. Phase one live inside a quarter — an ESM Foundation front door in weeks.
Incident, problem, change and request, built for deflection and resolution time, not ceremony.
Controls, policies and risk registers that regulators and auditors recognise.
Security Incident and Vulnerability Response wired to your SOC reality.
Indicators tied to the standard we agreed — measured, not decorated.
Connected enterprise data feeding workflows and AI without another migration.
Cases, workspaces and portals that hold up at your real volumes.
Work orders, scheduling and mobile field work on the same data model.
Demand, projects and portfolio decisions leadership will actually use.
ServiceNow ships AI governance out-of-box: AI Control Tower discovers, approves and logs the agents changing your platform. That's genuinely useful — and it's still a tool. A tool doesn't carry your accountability. Someone has to decide which policies it enforces, prove they match your obligations, own it once it's live, and sign to the auditor that it holds. That accountability is yours. It doesn't move to the platform, and it doesn't move to us. What we do is make the tool answerable to you.
When an autonomous change breaks a control, you are still liable. You can't tell a regulator the AI did it — or that the tool approved it. Accountability doesn't transfer to the machine, or to the machine that governs it.
We use AI in our own delivery too — to assess, document and test faster. It's a tool in how we work, never the thing that decides. A senior architect decides, and signs.
of Icelandic companies use AI software, but only 14% have an AI strategy. The gap between adoption and governance is the problem we're built for, and it looks the same across the Nordics.
Statistics Iceland · ICT 2024, published March 2026No pyramid, no bench — capacity is added only when a mandate needs it.
We don't resell licences and we don't take delivery volume targets. Our advice on scope, architecture and vendors has exactly one client: you.
Headquartered in Reykjavík, working across Iceland, the Nordics and Europe — in English, Icelandic and plain words.
A 30-minute conversation with a founder — about your instance or the one you're planning, your licences, and the number you're accountable for. If it's a fit, you'll get a scoped, fixed-price proposal with the success criteria written in.