Founding Engineer, Agent Systems

• Develop secure sandboxing for agent execution
• Implement prompt-injection defenses and other agent-security controls
• Build production evaluation frameworks for fuzzy, high-stakes outputs
• Curate evaluation datasets and regression suites
• Design and improve LLM-as-judge systems
• Evaluate model performance across upgrades and provider changes
• Build retries, fallbacks, and circuit-breaker mechanisms
• Own prompt versioning and AI-quality infrastructure
• Design systems around async processing, queues, and idempotency
• Define the internal standard for production AI quality
• Own AI reliability and make ship/no-ship decisions when necessary, * Node.js
• React
• Tailwind
• Express
• Azure
• Postgres
• Terraform
• GitHub Actions
• Docker
• Anthropic-first AI stack
• Claude Code

What you’ll get

• Founding-level ownership of the agent platform
• Significant influence over AI architecture and engineering standards
• Opportunity to build production-grade agent systems from an early stage
• Direct impact on customer-facing AI quality and reliability
• Exposure to frontier LLM APIs in demanding production environments
• Work on technically difficult evaluation and reliability problems
• Close collaboration with company leadership
• Top-decile London market compensation
• Meaningful EMI-eligible equity
• Significant AI tooling and API budgets
• Daily team lunch
• Specialty coffee
• King’s Cross office with roof terrace and on-site facilities
• Fast-moving engineering environment with high autonomy

• Strong professional backend engineering experience
• 1-2+ years of experience shipping production LLM-powered features
• Strong TypeScript or Node.js experience
• Experience building customer-facing production AI systems
• Hands-on experience with agent frameworks
• Experience with tool calling and multi-step orchestration
• Production experience building AI evaluations
• Experience with dataset curation and regression testing
• Understanding of LLM-as-judge approaches and failure modes
• Strong systems-thinking ability
• Experience with asynchronous systems and queues
• Understanding of idempotency and reliability patterns
• Strong product-engineering mindset
• Ability to own features from customer problem through production deployment
• High level of autonomy and accountability
• Based in London or able to commute regularly for an on-site-first role

Particularly valuable experience

• Anthropic APIs
• OpenAI APIs
• Open-weight models
• Production LLM workloads at scale
• Agent orchestration frameworks
• AI evaluation infrastructure
• Model fallback architecture
• Prompt versioning
• Prompt-injection defense
• Agent security
• Sandboxed tool execution
• Production RAG systems
• High-stakes AI applications
• Compliance
• Governance, Risk and Compliance (GRC)
• Audit technology
• Cybersecurity
• Financial services
• Regulated enterprise software

Strong profile signals

• Former founder or technical founder
• Former CTO returning to hands-on engineering
• Founding or early-stage engineering experience
• Experience at a high-growth product company
• Experience at a leading technology or AI organization
• Significant independent AI projects or side projects
• Strong evidence of product ownership
• History of shipping customer-facing systems end-to-end
• AI-native development workflow
• Heavy use of coding agents such as Claude Code or equivalent
• Ability to move between product, backend, AI systems, and reliability work
• Strong technical judgment rather than purely research-oriented expertise

This role may not be the right fit if you

• Have exclusively front-end or front-end-first engineering experience
• Have a purely AI research background without production software engineering ownership
• Have worked primarily in Developer Relations or Developer Experience
• Have only built internal AI tools rather than customer-facing production systems
• Lack meaningful backend engineering experience
• Have AI experience limited to prototypes or demos
• Prefer research over shipping production products
• Lack experience taking ownership of production reliability
• Prefer narrowly scoped engineering responsibilities
• Are uncomfortable making judgment calls around AI quality and production readiness
• Are not comfortable working primarily on-site in London

Technical environment

• Join a seed-stage company building AI-native risk infrastructure for regulated enterprises
• Work on production AI systems used across financial services, regulated technology, and healthcare
• Own the agent platform powering customer-facing AI features
• Build the orchestration, evaluation, and reliability layers around frontier LLMs
• Turn model capabilities into production systems that domain experts can trust
• Set engineering standards for when AI features are reliable enough to ship
• Operate with founding-team levels of ownership and technical influence
• Work in a company already generating seven-figure revenue through multi-year enterprise contracts

This is a role that TechTree is recruiting for on behalf of one of its clients.

TechTree is an AI-driven recruitment platform working with high-growth companies.