Forward Deployed Engineer | AI Agents | London

Space Executive; Greater London; Hybrid; GBP 120,000 - 150,000; Full time

Benefits offered by this job

Equity

Job summary

Space Executive is seeking a Senior Engineer in London to build AI agent systems for enterprise customers. You will split time between hands-on engineering and collaborating with customer teams, contributing to production deployments and monitoring.

You’ll work with the CTO and help shape product direction as the London team grows, with hybrid in-office requirements in central London.

Qualifications

• Proficiency in Python with production software shipped.
• Experience deploying LLM agents in production environments.

Responsibilities

• Build agent systems for customer projects, including code, tests and monitoring.
• Evaluate agent prompts, tools, model choices, and limits; make safe changes via configuration files.

Skills

Python

LLM agents in production

Tool use

Retrieval

Evaluations

Debugging model behaviour

Cross-time-zone collaboration

Senior/Principal level

Stakeholder communication

Job description

I’m working with a Series A, VC-backed AI company that builds agent systems for enterprise customers. They’re looking for a senior engineer in London to build those systems alongside customer teams and take them to production.

The role

You’d split your time roughly 70% hands‑on engineering and 30% customer work. You’d join a customer's team for a defined project, learn how their data, tools and users fit together, and then build the agent system yourself: the code, the tests and the monitoring.

A typical problem is an agent that decides for itself which source to search, whether documents, internal notes or a data API. It works in demos, then in daily use it answers some figures from memory, writes poor search queries and makes too many calls on broad questions. You’d find the cause of each fault, build evaluations that score the agent’s source choices as well as its final answers, and show that each fix doesn’t break earlier cases. You’d also move the agent’s prompts, tools, model choice and limits into configuration files, so the customer's teams can change its behaviour without a redeploy. That means building the checks, version history and rollback that make those changes safe.

On the customer side, you’d agree with their team what “correct” means and what error rate they can accept, review wrong answers with their domain experts, and report progress each week with numbers. When a problem needs a different approach, you’d be the one to say so.

You’d have a direct line to the CTO, feeding back what you learn from customers and helping decide what the company builds next, ideally before customers start asking for it. As the London team grows, you’d help hire it.

What we’re looking for

• Strong coding skills, mostly in Python, with production software shipped
• Experience with LLM agents in production, including tool use and retrieval
• Experience building evaluations for agents or other multi-step systems
• A track record of debugging model behaviour and finding the cause
• Comfort working inside a customer’s team and explaining technical trade-offs to people who aren’t engineers
• Senior or Principal level, or equivalent
• Used to working with colleagues across time zones

Useful but not essential

• Experience with configuration systems, feature flags or staged rollouts
• Background in a regulated sector such as finance or healthcare
• Time at an AI research organisation or a vendor to one
• Experience helping establish a new regional team

Package

• £120k to £150k base, plus equity.
• Hybrid in London, with 2 days a week in the office. Any travel would be to customer sites within central London and paid for.