What does an AI consultancy for UK financial services actually do?
We audit where AI fits inside your permissions and where it does not, build on the stack you already run (Microsoft 365, your CRM, your platform), and leave a senior engineer in the team to keep the work compliant under Consumer Duty, SMCR and the post-DUAA automated-decision regime. We do not sell strategy decks. The output is an auditable workflow that a senior manager can defend to the FCA.
Will AI replace advisors, paraplanners or compliance reviewers?
No. The FCA's position is that senior managers retain personal accountability under SMCR regardless of how a decision was reached. The DUAA 2025 requires meaningful human intervention by a reviewer with authority and competence to override the AI for any significant decision. We build copilots that let advisors handle more cases at higher consistency, with a citation-anchored audit trail per interaction.
How does this work with the FCA Handbook, Consumer Duty and SMCR?
The audit produces an SMCR ownership grid that names the senior manager accountable for each AI-assisted workflow, a Consumer Duty fair-value and foreseeable-harm review of every workflow where AI touches a customer outcome, and an evidence plan that documents model selection, testing and ongoing monitoring. The FCA confirmed in 2026 that it will not introduce AI-specific rules.
What about PRA SS1/23 if we are a bank or PRA-designated firm?
SS1/23 is technology-agnostic but explicitly captures AI and ML models, and the PRA flagged AI adoption as a 2026 supervisory priority. For PRA-regulated firms we extend the audit to include the SS1/23 five principles and produce model-risk documentation that fits inside your existing MRM framework.
Can we keep client data on-prem or in a UK / EU controlled environment?
Yes. Anthropic Claude on UK / EU regions or OpenAI on Azure UK / EU as the default cloud options, open-weights models on private GPU for sovereign requirements, and full on-prem for firms where data residency rules out cloud LLMs entirely. We also ship zero-retention OCR. Every deployment carries an ISO 27001-aligned audit logging trail and a UK GDPR data processing register.
Does this integrate with our CRM, platform and Microsoft 365 estate?
Our verified integrations include Microsoft 365 (SharePoint, document libraries, Entra ID), Power Automate for workflow integration, Salesforce, REST/SOAP API connections, on-prem PostgreSQL, and Langfuse for prompt versioning and trace logging. We do not name core-banking platforms we have not delivered against; we scope the integration honestly during the audit.
Who is liable if an AI-assisted decision goes wrong under Consumer Duty?
Liability sits where it has always sat under SMCR: with the named senior manager accountable for the function. The DUAA 2025 reinforces this — meaningful human intervention is a statutory safeguard, and a human rubber-stamping the AI output does not meet it. We design the workflow so the human reviewer has the citations, the retrieved context, the model's confidence signal, and a real path to override.
How long does a financial services AI engagement take?
The audit is fixed-fee and fixed-scope. A Transformation Block is a build-and-train engagement priced per scope and ships an integrated pilot inside your existing stack with the team trained on it. The Senior AI Lead is a rolling monthly retainer that begins once you have a Transformation Block in production. A bespoke build (the Kaption pattern) is scoped after the audit. Numbers are shared during a discovery call so we can size against what you actually want to do.
What does it cost?
Pricing is fixed before kickoff and bands are shared on a discovery call. The audit is fixed-fee, the Transformation Block is priced per scope, the Senior AI Lead is a rolling monthly retainer, and bespoke builds are quoted after the audit so the price reflects what the audit actually surfaced. No discovery-call surprises on the invoice. No retainer required to access the audit.
How is OpenKit different from the Big Four or a fractional CAIO consultancy?
The Big Four split strategy and implementation across separate teams on separate budgets; the senior engineer on day one is rarely the senior engineer on day ninety. Faculty AI (now part of Accenture) and the larger Gemini-set firms — Digiterre, Scott Logic, BJSS, Altus Consulting — operate at enterprise scale with day-rate or T&M engagements. The fractional CAIO retainer puts an advisor in the room but does not write code. The productised offerings — Wealth IQ, Dailoqa, Holistic AI — sell a product, not a build against your own corpus. We do both: a senior engineer audits the work, scopes the build, ships the integration, trains the team, and stays on as the Senior AI Lead. Our audit is fixed-fee, which the buyer can self-qualify against without a procurement cycle.
