# Hire an AI engineer, or an implementation partner? > Choosing an AI implementation partner in the UK or hiring an AI engineer: real UK costs, what to ask before you sign, who owns the code, and what we build. Source: https://www.assume-ai.com/ai-implementation-partner-uk An AI implementation partner designs, builds and deploys working AI systems inside your firm, then hands them over. For a UK firm of 20 to 200 people the alternative is hiring an AI engineer, at a median advertised salary of £87,500, or about £101,200 a year once employer National Insurance and pension are added. AssumeAI is a UK, founder-led AI consultancy and AI engineering firm that builds for consultancies and professional services firms at published prices, and you own everything we build, code included. ## What does an AI implementation partner actually do? An AI implementation partner takes a business problem and leaves you with a working system: an automation, an assistant, an agent or a platform your people use every day. The work runs from deciding what is worth building, through governance and the build itself, to testing with the people who will use it, deployment into your existing tools and a handover your team can run without the partner. That is different from two neighbours it is often confused with. A strategy firm tells you where AI could help and stops at the report. A software house builds to a specification but will not tell you whether the specification is the right one. AssumeAI sits between the strategy firms that cannot build and the AI development firms that cannot advise, so the same team that agrees what to build is the team that builds it. For a consultancy or professional services firm of 20 to 200 people, the honest question is not whether to use AI but who will do the building. The barrier is rarely the money; it is not having anyone in house who can build. There are three ways to close that gap: hire an AI engineer, engage a contractor, or bring in a partner. Each has a place, and the rest of this guide sets out what each one really costs and when it fits. ## What does it cost to hire an AI engineer in the UK? More than the salary on the advert. In the six months to 21 October 2026, the median advertised salary for an AI engineer in the UK was £87,500, from 504 salaries quoted, with the 10th percentile at £53,750 and the 90th at £115,000 (ITJobsWatch). Advertised salaries are not the same as agreed pay, but they are a fair guide to what you will be competing against. On top of the salary sit the costs every employer carries. For the 2026 to 2027 tax year, employer National Insurance is 15% on earnings above £5,000 a year (GOV.UK), and under automatic enrolment you must pay at least 3% of qualifying earnings, between £6,240 and £50,270, into a workplace pension (GOV.UK). For an engineer on £87,500 that works out as follows: - Salary: £87,500 - Employer National Insurance: 15% of £82,500, which is £12,375 - Minimum employer pension: 3% of £44,030, which is about £1,321 - Total: about £101,200 a year, or roughly £8,400 a month That is before a recruitment fee, a laptop and software, the paid models and cloud services the engineer will need, holiday and sickness, and the time a manager spends directing someone whose work they may not be able to review. If your firm qualifies for the Employment Allowance, which is £10,500 for 2026 to 2027, it reduces the National Insurance across all your staff, not this one hire. The harder cost is what one person cannot cover. A useful AI system in a professional firm needs someone who can find the right problem, write the governance, integrate with your practice management and finance systems, build, test with sceptical users, document and train. Hiring one engineer asks one person to be good at all of it, with nobody to check their work and nobody to cover when they leave. ## Partner, hire or contractor: which fits your firm? Each route suits a different firm. Here is how we would think about it if we were in your seat. ### Hire an AI engineer Hire when AI is going to be a permanent, central part of your product or service, you have enough work to keep an engineer busy for years, and someone senior in the firm can direct and review technical work. A permanent hire builds knowledge that stays in the business. The risk is a long, expensive search and a single point of failure once they arrive. ### Engage a contractor Engage a contractor when you know exactly what you want built and can specify it well. The median UK contract rate for an AI engineer was £600 a day in the six months to 21 October 2026 (ITJobsWatch), which is about £12,000 for a twenty-day month. Contractors are quick to start, but most will build to your brief rather than question it, and the off-payroll working rules (IR35) may mean your firm decides their employment status if it is medium or large (GOV.UK). ### Bring in a partner Bring in a partner when you are not yet sure what to build, when governance and client data matter, or when you want working systems quickly without carrying a hire. A good partner brings a method, a team and a record of builds in firms like yours, and leaves capability behind. The risk is lock-in, which is why the next section matters. ## What should you ask an AI implementation partner before you sign? Most of the risk in choosing an AI implementation partner can be found with a handful of plain questions. Ask every firm on your shortlist, including us. - Who owns the code, the prompts, the configuration and the data when the work ends? The answer should be you, in writing. - Where will our client data be held and processed, and which policies will it be handled to? - What governance comes before the build: an AI policy, impact assessments, a risk register? - Will you tell us what not to build? Ask for an example of something they talked a client out of. - How do you decide what to build first, and what return is that decision based on? - Who will actually do the work, and what happens if that person is unavailable? - What does the handover include: documentation, training, walkthroughs? - Can we see builds you have delivered for firms like ours, with figures the client collected? - What is the price, what does it include, and what happens if the build is not working on time? A partner worth hiring will answer all of these without hesitation. Be wary of anyone who quotes a build before they know where your hours go, or whose answer to ownership is a licence to use what they built. ## What does AssumeAI build, and what does it cost? Four ways to work with us on implementation, each with a published price and its own guarantee. ### 30-Day Implementation Sprint The [30-Day Implementation Sprint](https://www.assume-ai.com/services/30-day-sprint), from £6,000, delivers one use case done properly: one week to understand your pain points and agree what to build, three weeks rolling it out, integrated with the tools you already use. If it is not live and delivering value in 30 days, we keep working at no extra cost until it is. ### 90-Day Implementation Sprint The [90-Day Implementation Sprint](https://www.assume-ai.com/services/90-day-sprint), from £15,000, is the full arc: two weeks of stakeholder interviews and pain point workshops, eight weeks developing and deploying the solutions that make the biggest difference, and two weeks embedding the capability in your team. We agree the outcomes up front, and if we do not hit them we keep going until we do. ### Enterprise AI Deployment [Enterprise AI Deployment](https://www.assume-ai.com/services/enterprise-ai-deployment), at £20,000 excluding VAT, puts Claude Cowork or ChatGPT Enterprise on EU-based data hosting, configured to your data governance policies, with six to twelve agents and skills built around your real workflows. It is designed for firms of 20 to 200 employees, and if you are not happy, for any reason, we refund you in full. ### Managed Service The [Managed Service](https://www.assume-ai.com/services/managed-service) embeds half an engineer from £5,500 a month, one from £8,500 or two from £15,000, side by side with your people, rolling out your roadmap and transferring capability as they go. It is the closest thing to hiring an AI engineer without the overhead of hiring one, on a monthly footing you can scale up or down. ## Who owns the code, and what happens when the work ends? You do. We do not build on top of your business, we build inside it. Everything lives in your existing infrastructure and you own all of it, code included. We build around your processes rather than around one model, so as the tools improve you can change what sits underneath without starting again. Every build ends with a handover designed so that you no longer need us for the work we did. At [Coaching Without Borders](https://www.assume-ai.com/work/coaching-without-borders-transformation-programme) that meant a start-here page, an operator guide and walkthrough films for each automation. Enterprise deployments come with documentation of every agent and skill, so your team can maintain and extend them. The 90-day sprint ends with two weeks spent embedding the capability in your team. We would rather you came back because the first build paid off than because you could not leave. Benj Johns, CFO of Sladen Consulting, described working with us as “a partnership process, rather than a supplier and client process”. - The code, which you own - A start-here page and an operator guide - Walkthrough films for each automation - Documentation of every agent and skill - Governance your team can run itself ## What has AI development delivered for consultancies? The test of an implementation partner is what is still running after they leave. These are builds for consultancies and professional services firms, with figures collected from the people who use them. At [The Litmus Partnership](https://www.assume-ai.com/work/litmus-partnership-transformation-programme) an eight-month programme of AI and automation now saves 268 hours a month. Its benchmarking platform handles 2,000 lines in around 45 minutes, where 200 to 300 lines used to take eight hours. Katie Foster, on film for Litmus, put it this way: “I feel like I’ve gained a full-time member of staff, and it’s not costing any extra money.” For [Clear AIMS](https://www.assume-ai.com/work/clear-aims-proposal-generator) we built a proposal generator that takes the average proposal from two to three hours of work down to five to ten minutes. At [Coaching Without Borders](https://www.assume-ai.com/work/coaching-without-borders-transformation-programme), one form now sets up a new project in four systems in about ninety seconds. [The Consultancy Growth Network](https://www.assume-ai.com/work/consultancy-growth-network-ai-board-advisor) has an AI board advisor trained on more than 400 hours of its own material, available to every member. You can see every project, with the figures, on our [work page](https://www.assume-ai.com/work). If you run a consulting firm, our guide to [AI for consultancies](https://www.assume-ai.com/ai-for-consultancies) covers where AI pays back first. ## Who would actually build it? AssumeAI is founder-led from end to end. [Hugh Dowling](https://www.assume-ai.com/about#hugh), Head of Development, completed a Master's in Software Engineering and makes sure everything we build is designed to operate in a secure, scalable, commercial environment. [Rupert Wiseman](https://www.assume-ai.com/about#rupert), Managing Director and Principal AI Consultant, studied Law at Exeter and worked in corporate finance, and his legal background brings a rigorous focus on safety, governance and risk controls from day one. Our AI engineers and consultants work alongside them, and behind the founders is an associate network of over ten AI consultants and AI engineers, so the work is never waiting on one person. That is the practical difference from a single hire. Meet the team on our [about page](https://www.assume-ai.com/about), and see how governance fits around a build in our guide to [AI for professional services](https://www.assume-ai.com/ai-for-professional-services). ## Where do the market figures come from? The salary, day rate, tax and pension figures on this page were checked on 21 October 2026. Advertised salaries and day rates move from month to month, so treat them as a guide. - [ITJobsWatch: Artificial Intelligence Engineer jobs, UK, permanent salaries, six months to 21 October 2026](https://www.itjobswatch.co.uk/jobs/uk/artificial%20intelligence%20engineer.do) - [ITJobsWatch: Artificial Intelligence Engineer contracts, UK, daily rates, six months to 21 October 2026](https://www.itjobswatch.co.uk/contracts/uk/artificial%20intelligence%20engineer.do) - [GOV.UK: Rates and thresholds for employers 2026 to 2027](https://www.gov.uk/guidance/rates-and-thresholds-for-employers-2026-to-2027) - [GOV.UK: Workplace pensions, what you, your employer and the government pay](https://www.gov.uk/workplace-pensions/what-you-your-employer-and-the-government-pay) - [GOV.UK: Understanding off-payroll working (IR35)](https://www.gov.uk/guidance/understanding-off-payroll-working-ir35) ## Questions ### Is it cheaper to hire an AI engineer or use an AI implementation partner? On salary alone the two are closer than most firms expect. The median advertised salary for an AI engineer in the UK was £87,500 in the six months to 21 October 2026 (ITJobsWatch), which comes to about £101,200 a year once employer National Insurance and the minimum pension contribution are added, or roughly £8,400 a month, before recruitment, equipment and management time. AssumeAI's managed service embeds one engineer from £8,500 a month, or half an engineer from £5,500, on a monthly footing. The real difference is what comes with it: a partner brings a team, governance and a method, while a single hire has to cover every skill alone. ### Can I hire a part-time AI engineer for my firm in the UK? Yes, in three ways. You can engage a contractor for a few days a week; the median UK contract rate for an AI engineer was £600 a day in the six months to 21 October 2026 (ITJobsWatch), and the off-payroll working rules (IR35) may apply. You can employ someone part time. Or you can use a managed service: AssumeAI embeds half an engineer inside your firm from £5,500 a month, with the founders and an associate network of over ten AI consultants and engineers behind them, and you can change the level or step away on a monthly footing. ### Who owns the code an AI implementation partner builds? It should be you, and you should get that in writing before anything is built. With AssumeAI you own all of it, code included. We build inside your existing infrastructure rather than on top of it, and every build ends with a handover: documentation, and for automations an operator guide and walkthrough films, so you are never locked in to us. ### What can an AI implementation partner build in 30 days? A focused, genuinely useful solution for one clear workflow: an automation, an assistant or an agent. AssumeAI's 30-day implementation sprint spends one week agreeing what to build and three weeks rolling it out, from £6,000. If it is not live and delivering value in 30 days, we keep working at no extra cost until it is. ### Do we need an AI strategy before we hire an implementation partner? It helps, but it is not essential. If you already know the bottleneck you want gone, a 30-day sprint spends its first week agreeing exactly what to build, and a 90-day sprint spends its first two weeks on stakeholder interviews and pain point workshops. If you are not sure where AI would pay off, an AI Strategy Review or an AI Priorities Workshop will stop you spending on the wrong things.