Prompt engineers design, test, and optimise the text inputs used to get useful, accurate, and safe outputs from large language models like GPT-4, Claude, or Gemini. The job sits at the intersection of linguistics, UX, and machine learning — you're essentially writing instructions that reliably shape model behaviour at scale. Most roles are inside AI product teams, working closely with software engineers, data scientists, and product managers. A typical week involves iterating on system prompts, running evaluation datasets, documenting prompt failures, and writing internal guidelines for how the team interacts with AI systems.
Sam Blackwell
AI Prompt Engineer
📍 London, UK✉️ sam.blackwell@email.com
Summary
AI Prompt Engineer with 3 years of experience designing and optimising LLM-powered applications. Expert in prompt engineering, RAG architectures, and AI safety evaluation. Background in computational linguistics.
Work Experience
Prompt Engineer at AnthropicJun 2023 — Present
Design and evaluate system prompts for enterprise AI assistant products serving 50K+ daily users
Build evaluation frameworks measuring response quality, safety, and helpfulness across 10K+ test cases
AI Research Engineer at Stability AIAug 2021 — May 2023
Optimised text-to-image prompts improving CLIP score by 15% for Stable Diffusion models
Built automated prompt testing pipeline evaluating 100K+ prompt variations per experiment
Prompt Engineer CVs are relatively new but increasingly in demand. Recruiters want to see experience with production LLM systems, evaluation methodologies, RAG implementations, and a background in linguistics or ML.
Key Skills to Include
Prompt engineering, RAG architecture, LLM evaluation, Python, LangChain, AI safety, fine-tuning, NLP, structured output design, and red teaming.
Common Mistakes
Underselling the technical depth of the role. Prompt engineering in production is much more than writing prompts. Show your evaluation frameworks, accuracy improvements, and system design contributions.
Formatting Tips
One to two pages. Include a Projects section describing production systems. Link to publications, blog posts, or open-source contributions.
Average Salary — Prompt Engineer
United States
$95,000 – $160,000
United Kingdom
$60,000 – $100,000
Germany
$65,000 – $105,000
UAE / Dubai
$70,000 – $115,000
Canada
$80,000 – $130,000
Australia
$85,000 – $135,000
Figures in USD. Ranges reflect mid-level experience (3–7 years). Senior roles and major metro areas typically sit at the top of these bands.
Top 5 Interview Questions — Prompt Engineer
1How would you improve a prompt that is giving inconsistent outputs across different runs?
Talk about adding explicit constraints, few-shot examples, chain-of-thought instructions, and output format specifications. Show you approach prompt debugging systematically rather than intuitively.
2How do you evaluate whether a prompt is performing well?
Describe building an evaluation dataset, defining clear success criteria, running A/B comparisons, and using automated scoring alongside human review. Good prompt engineering is empirical, not subjective.
3What's the difference between zero-shot, few-shot, and chain-of-thought prompting?
Zero-shot asks the model with no examples. Few-shot provides worked examples in the prompt. Chain-of-thought asks the model to reason step-by-step before answering. Be ready to say when you'd choose each and why.
4Tell me about a prompt you built that had to handle edge cases or adversarial inputs.
Describe the use case, what edge cases you anticipated, how you built in guardrails, and how you tested robustness. This tests both technical depth and product thinking.
5How do you stay current with developments in LLM capabilities and prompting techniques?
Name specific resources: Anthropic's prompt engineering guide, OpenAI cookbook, arXiv papers on reasoning and alignment, the DAIR.AI prompting guide. Specificity signals genuine engagement with the field.
How to Tailor Your CV
Anthropic, OpenAI, Google DeepMind, and enterprise AI teams at companies like Salesforce and Microsoft are the highest-profile employers. Pure AI labs want to see technical depth — familiarity with model evaluation, red-teaming, and alignment concepts alongside the prompting craft. Enterprise companies want a more product-focused profile: evidence you've shipped AI features at scale and can work across product, engineering, and business stakeholders.