LEC AI
Lead AI engineer - Agent Systems
Get notified of more jobs like this · No spam, ever
How your CV stacks up
Upload your CV to see how well it fits this job role
?%
Lead AI Engineer — Agent Systems
LEC AI · London · Full-time · Office-based
Location and commitment:
This role is based in our London office five days a week. During major system launches, additional effort — including occasional weekends — is expected. We state this upfront so expectations are clear from the first conversation. Good salary for the top engineers.
About LEC AI
LEC AI is the artificial intelligence business within LEC Industries, a London-headquartered group building technology and operating companies across robotics, AI and consumer brands, with seven decades of operating history between the UK and China. We are a technology business with live commercial deployments and real users, not a research lab or a corporate IT function.
The role
We are hiring a Lead AI Engineer to own the agent architecture across three connected production systems:
- A marketing operating system in live production
- An editing intelligence trained on documented human judgement
- A large-scale autonomous agent platform of 34 specified agents that researches, builds, tests and grades its own work, with human oversight at defined boundary points
All three run on a self-hosted stack we own end to end: vector and graph memory, an audio processing pipeline, and model-agnostic serving across hosted and local models. You will be the architect of how these agents coordinate, hand off work, fail safely, recover and improve.
Why this role is distinctive
- Greenfield at scale. The 34-agent platform is fully specified and not yet built. You will build it from the first file and own it.
- Outcomes-driven. Work is measured and graded on results, with fast decision-making rather than lengthy roadmap processes.
- Senior access. You report to the Principal AI Architect and work directly with the Group CEO. Decisions are made in hours, not quarters.
- Frontier problems. Multi-agent handoff without information loss, agents that safely revise their own operating methods, ensemble judgement systems, containment procedures for misbehaving agents, and cost governance for autonomous loops.
- Full ownership of the stack. Self-hosted infrastructure, with the freedom to choose and change the tooling.
Reasons to use Rodeo
Start with a chat, not a search bar
Grad scheme, placement, apprenticeship? Not sure what you want yet — that's fine. Your agent talks it through with you and turns "I have no idea" into a shortlist.
It searches the market for you
Every day your agent scans the market matching roles against what actually matters to you, not just keywords on a CV.
Only hits
No noise. No "maybe this fits." Just roles with a clear explanation of why they're right — and where to focus when applying.
Key responsibilities
- Design and build the full agent roster: squads for intelligence gathering, verification, experiment design, launch and grading, coordinated through a message bus with a single-writer rule and daily reconciliation
- Build multi-agent judgement panels — groups of specialist agents that assess work independently, exchange views and record a chaired verdict, with all reasoning logged as training data
- Implement agents as versioned, file-defined systems capable of safely revising their own methods under controlled conditions
- Design the boundary layer: full autonomy within the system, with a defined set of actions that always require human approval before taking effect externally
- Establish reliability engineering across the agent estate: retries, idempotent operations, reconciliation and failure recovery
Get help with your application
Your very own career expert that helps elevate your application to the next level.
What we are looking for
- Demonstrated delivery of a production multi-agent system with real users — able to explain its architecture, handoff design and failure modes in detail
- Practical experience of the failure modes of agentic systems — runaway loops, lost handoffs, cost overruns, unsafe retries — and the engineering patterns that prevent them
- Sound judgement in framework selection, with working fluency in the current open-source agent ecosystem; familiarity with leading Chinese open-source frameworks (e.g. AgentScope, Qwen-Agent) is a distinct advantage, as we track this ecosystem as a discipline
- A builder’s profile: experience of small teams, end-to-end ownership and rapid delivery, with achievements evidenced by what you have personally built and shipped
Who this is for
We are looking for hungry people. Builders who want to own something and make it work, who move fast because they can’t help it, and who take pride in what they ship. If you are looking for a comfortable seat, a big-company routine or a role you can coast in, this is not it. If you want to build the most ambitious agent system you’ve ever been given the keys to, it is.
How to apply
Email talent@lecai.ai with your CV, links to work you have built, and anything further you want to tell us — the more evidence of what you have personally shipped, the better. We aim to complete the full process, including a practical exercise, within two weeks.
“It took my CV and asked me questions relevant to understanding what kind of jobs to suggest for me. Suggestions were almost perfect. Jobs were exactly what I’ve been looking for.”
Jessica, London
Skills
Location