Applied LLM Engineer

Role - Applied LLM Engineer

Salary - Up to £100k + Equity

Location - Edinburgh OR London (Hybrid - Min 3 days in office)

Wordsmith

Most legal teams are drowning. They’re buried under internal questions, contract reviews, policy approvals, and fire drills from every corner of the business.

Wordsmith is the AI command center for in-house legal. We automate the chaos—intake, Q&A, redlines, drafting, and research—so legal can finally operate at the speed of business.

Backed by Index Ventures and some of the sharpest minds in law and AI, we’re scaling fast across London, New York, and beyond. Our customers include fast-growth tech companies and public enterprises. We’re building the future of legal work.

The Role

Join us in redefining how legal teams work by building state-of-the-art AI-native experiences powered by Large Language Models.

As an AI Engineer, you’ll work across backend systems, evaluation frameworks, and agentic LLM workflows to create scalable, secure, and continuously improving AI features. You’ll partner closely with design, product, and legal teams to bring thoughtful, high-impact functionality to life.

What you'll do

• Build and integrate LLM-powered features in production.

• Define success metrics (task success, grounding, latency, cost) and set practical SLOs.

• Create and maintain evaluation datasets: golden sets, adversarial cases, and regression suites.

• Automate quality checks with an offline eval harness, CI gates, and safe online tests (A/B, canaries).

• Instrument tracing and logs for prompts, responses, errors, and costs; analyze failures and ship fixes.

• Improve quality end-to-end: prompt design, retrieval (RAG), model selection/fine-tuning, and post-processing.

• Work across APIs, data pipelines, and backend services to deliver complete user workflows.

• Collaborate with product and design so the AI feels helpful and trustworthy.

• Keep privacy, reliability, and performance front and center.

Experience & Qualifications:

• 4+ years building user-facing software (B2B SaaS is a plus).

• Strong Python skills and comfort with at least one other backend language.

• Production experience building with LLMs or similar AI services.

• Solid understanding of APIs, cloud infrastructure, and scalable backend design.

• Proven loop for improving model quality:

  • Building/maintaining eval datasets and regression suites

  • Running offline evals and online experiments, interpreting results

  • Observability for LLM systems (traces, latency/cost tracking, drift detection)

• Product sense and curiosity—you think in terms of user outcomes, not just code paths.

What you can expect

• A high-trust, high-impact environment with ownership over mission-critical AI systems

• Challenging product problems at the intersection of law, language, and intelligence

• A collaborative team of engineers, designers, and legal experts building something new

• Competitive salary, meaningful equity, and a say in shaping both product and culture

If you’re looking for more than just a job - a place to learn, grow, and have a real impact from day one - this could be the one. We’re building something exciting, and we’re looking for people who want to be part of the journey. If that sounds like you, hit apply.