How to Become an AI Prompt Engineer in the UK
AI prompt engineers design and test the instructions, examples and workflows that make large language models do useful work in real products. Day to day that means writing system prompts, building test sets and evaluations, spotting where a model fails, and adding guardrails, retrieval or tool calls to fix it. In many companies the work sits inside a wider AI engineering or product team.
The field is new, so there is no standard route or recognised qualification. Most people move across from software development, data, technical writing or conversation design, and prove themselves with working projects and evaluation results.
Quick facts
- Starting salary
- £35,000 - £50,000
- Mid-career salary
- £55,000 - £80,000
- Senior salary
- £90,000 - £140,000+
- Work environment
- Remote, hybrid, fast-moving teams
- Time to entry
- 6 - 18 months
- Degree required
- Preferred - portfolio matters more
- Category
- AI, Data and Automation
Compare pay across careers in the UK salaries by career report. Also in: High-paying careers, No-degree careers, Creative careers, Career change.
What a AI Prompt Engineer does
Design, test and refine prompts and pipelines that get the most out of LLMs in production - eval, guardrails and fine-tuning included.
- Applied Prompt Engineer (SaaS) - Build and maintain LLM features inside production apps.
- AI Eval Specialist - Design evaluation frameworks, golden sets and regression tests for LLM features.
- Conversation Designer - Design conversational flows, personas and guardrails for AI products.
How to become an AI prompt engineer in the UK
There is no required degree, licence or professional body for prompt engineering in the UK, and job titles vary: the same work may be advertised as AI engineer, LLM engineer, applied AI specialist or conversation designer. Employers mostly want Python, experience with model APIs, the ability to design evaluations and measure results, and clear writing. A degree in computer science, maths, linguistics or a related subject helps, and software or data apprenticeships and Skills Bootcamps can give you the coding foundations. Be wary of courses promising a recognised prompt engineering certification, because employers do not rely on any single one.
- Learn Python well enough to call model APIs, handle JSON and run batches of test prompts from a script.
- Pick one real task, such as summarising support tickets, and write a prompt, a test set of examples and a way to score the outputs.
- Measure how changes to your prompt, examples or model choice affect the scores, and write up what improved and why.
- Build two or three portfolio projects that use retrieval, tool calls or structured output, and publish the code and evaluation results on GitHub.
- Read the official prompting and evaluation guides from the main model providers, and update your projects as models and features change.
- Apply for AI engineer, LLM engineer, AI evaluation and conversation designer roles as well as prompt engineer roles, because titles vary widely.
- If you already work in software, support or content, offer to build an LLM feature or internal tool at your current employer.
| Route | How long | Cost | Best for |
|---|---|---|---|
| Move across from a software or data role | 6-18 months | Low, mostly your time | Developers and data analysts |
| Self-taught portfolio projects | 6-18 months | Low, plus API usage costs | Career changers who can code |
| Computer science or related degree | 3 years | Tuition fee loan | School leavers wanting a broad base |
| Conversation design or technical writing first | 1-2 years | None, salaried role | Strong writers with some technical skills |
How long it takes
Expect 6 to 18 months to build the coding, evaluation and portfolio work employers look for. If you already work as a developer or data analyst, an internal AI project can get you there faster.
AI Prompt Engineer salary in the UK
Pay for AI prompt engineering work usually starts around £35,000 to £50,000. Mid-level roles typically pay £55,000 to £80,000, and senior specialists can earn £90,000 to £140,000 or more. Pay depends heavily on how technical the role is: jobs involving Python, evaluation pipelines and production systems pay far more than roles focused on writing prompts for content. AI companies, large tech firms and London employers pay the most. Because the field is new, titles and pay are changing quickly.
Key skills
- LLM internals
- Prompt design + evals
- Python
- Clear writing
AI Prompt Engineer FAQs
How do I become an AI prompt engineer?
Learn Python and how to call model APIs, practise designing prompts with proper test sets and evaluations, and build a portfolio of working projects. Many people move across from software, data or content roles, often by building an AI feature in their current job.
Do I need a degree to be a prompt engineer?
No degree is required, but many employers prefer one in computer science, maths, linguistics or a related subject. A portfolio showing working projects and measured evaluation results usually matters more than the degree itself.
How much does an AI prompt engineer earn in the UK?
Starting pay is around £35,000 to £50,000, mid-level roles pay £55,000 to £80,000, and senior specialists can earn £90,000 to £140,000 or more. Technical roles involving Python and evaluation work pay the most.
Is prompt engineering a real long-term career?
The skills are in demand, but the standalone title may not last as models improve and the work merges into AI engineering and product roles. Building coding, evaluation and system design skills alongside prompting gives you a safer long-term path.
Is AI Prompt Engineer right for you?
Career Compass ranks careers from your interests and priorities and explains every suggestion, with UK routes and trade-offs. See how well AI Prompt Engineer actually fits you. £19.99 once, full report included.
Not sure AI prompt engineer is the right career for you? A good career test narrows the field before you spend money retraining.