AI consultancy · London · UK-wide

An AI consultancy in the UK that tells you the price first

Kelriva AI is a two-person London consultancy. We build document processing, workflow agents and analytics for UK companies at a fixed fee from £4,500, delivered in weeks, on your infrastructure, with the source code handed over at the end.

Book a 30-minute callSee the three fixed prices
Why this page exists

Most AI consulting in the UK is sold by the day, and the first invoice arrives before anything works

Search for an AI consultancy and you find two kinds of firm: large practices whose minimum engagement starts well above what a mid-sized company wants to risk on a first project, and agencies that sell strategy decks and workshops before any code exists.

We started Kelriva because we had sat on the buying side of that. So we publish our prices, scope in writing before we start, and measure ourselves on whether the system is running in your business at the end, not on hours billed.

We are small, and we say so. That is a feature for the size of project we take on and a limit for the ones we do not.

How we are different

Four things to check any UK AI consultancy against

Pricing

A fixed fee, agreed before we start

Three packages, three prices, published on the site: £4,500, £8,500 and £15,000. No day rates, no discovery phase that quietly becomes an invoice. If the scope changes, we re-scope in writing first.

Timing

Weeks, not quarters

The assessment takes about a week. A document-processing proof of concept or a working agent takes four to eight. You see running software early, not a slide deck at the end.

Ownership

Built on your infrastructure, then handed over

Source code, documentation and training are part of every package. The system is yours to run and extend, and nothing is tied to a subscription with us.

Honesty

A small team that says what it will not do

Two people in London. We take the work we can deliver ourselves, tell you when a task does not need AI at all, and never guarantee an accuracy figure before we have seen your documents.

What we build

Three things, done properly

Intelligent document processing

Extraction from contracts, KYC packs, compliance filings and invoices, with every field traceable back to the source page and a review step for anything the model is unsure about.

Details →

AI workflow automation

Multi-step LangGraph agents that replace manual approval chains and coordination work, with human-in-the-loop controls and a full audit trail.

Details →

Data analytics and intelligence

Pipelines, warehouses and dashboards that turn scattered operational data into something a manager can act on, including predictive views where the data supports them.

Details →
How an engagement runs

From a first call to a system your team runs

01

A conversation

Thirty minutes on what is still done by hand, what it costs, and what a working system would need to prove. If we are not the right fit, we say so on the call.

02

A written scope and a fixed price

You get a one-page scope with the deliverables, the timeline and the fee. Nothing starts until you have signed it.

03

Build in the open

Working software every week, on your infrastructure, with your team looking over our shoulder. Problems surface early instead of at handover.

04

Handover and a reference

Documentation, source code and training. Then, if it went well, we ask whether you would take a reference call from the next client.

Reference · Live in production

Coach matching at bettercoach: from seven days to seconds

bettercoach matched clients to coaches by hand. We built a privacy-aware matching engine that reads unstructured, multilingual enquiries and returns a ranked, explainable shortlist a human operator reviews. Their Managing Director described us as a “trustworthy partner” on “a topic of high complexity” and offers reference calls.

Read the case study
Discuss this with AI:
Ready when you are

Tell us what is still done by hand

One call is enough to scope it. You will have a written price within 48 hours, and a clear no if it is not a fit.

info@kelriva.ai
FAQ

Choosing an AI consultancy in the UK: what buyers ask us

What does an AI consultancy in the UK typically charge?

+

Most UK AI consultancies and agencies price by the day or by a retainer, and a first engagement of a few weeks often runs well into five figures before anything is proven. We publish three fixed prices instead: an AI Readiness Assessment at £4,500, an IDP Proof of Concept at £8,500, and a LangGraph Agent MVP at £15,000, each scoped in writing before work begins.

Are you based in London?

+

Yes. Kelriva AI Limited is registered in London and the two of us work with clients across the UK, mostly remotely, with on-site sessions where a project needs them.

How big is the team?

+

Two people, and we say so up front. It means the people you speak to on the first call are the people who build the system. It also means we take on a small number of engagements at a time and turn work away when we are full.

Do you have a live reference client?

+

Yes. bettercoach runs our AI-assisted coach matching engine in production. Their Managing Director has described the engagement publicly and is available for a reference call. The full case study is on this site.

Which industries do you work with?

+

Fintech and financial services, ESG and sustainability reporting, corporate finance and private equity, and corporate coaching and learning. The common thread is document-heavy, regulated work where a wrong answer has a cost.

What if we are not sure we are ready for AI?

+

Start with the AI Readiness Assessment. It scores six dimensions of your organisation, tells you honestly what to build first, and what not to build at all. If you then commission a build within thirty days, £1,000 of the assessment fee is credited against it.