**Please do not apply if you\'re not currently based in London. We are only accepting applications from those based locally.**
Quick Overview
This is a unique opportunity to join as a founding member of the team at a venture-backed startup.
We wanted to be transparent upfront about this role; we very much see you as a new addition to the founding team. This comes with significant workload and responsibility, but is also reflected in the very generous equity package available.
About Sawyer
There are billions of gamers out there. Yet many studios are still struggling. The market is competitive, dominated by established hits. Games are increasingly complex with ballooning budgets.
We believe that AI agents are the next big workplace shift. We believe that they are non-negotiable in today\'s world - and that every studio must utilise them to stay competitive.
To help studios win, we\'re building Sawyer - a SOTA agent created specifically for game studios. Sawyer sits at the heart of every studio\'s context and is capable of autonomously completing complex tasks for team members. Whether you\'re a programmer building new gameplay mechanics or a game designer reviewing the latest analytics, Sawyer is able to support you across these tasks.
We\'re working with some of the top studios in the world & are releasing a new and improved version of Sawyer in January.
We\'re also backed by some of the best investors in the UK & US and anticipate fundraising again in early 2026.
About the role
We\'re looking for an AI Engineer capable of building and improving SOTA agentic systems. You\'ll have autonomy in pushing the envelope of what is possible, helping us dictate the future of the industry.
The job responsibilities are as follows:
- Design, build, and maintain agentic AI systems (planners, loops, search algorithms, tool-using agents, etc.) for production use.
- Integrate MCPs & tools so agents can take real actions, not just generate text.
- Build orchestration and control flows (state machines, workflows, retries, monitoring) around LLM calls to make agents robust, safe, and observable.
- Implement context and memory systems (vector stores, search, scratchpads, summaries) so agents can operate over large contexts and long-running tasks.
- Ship agents as production services: containerize, deploy, and operate them (logging, metrics, tracing, alerting, performance tuning).
About you
- Strong software engineering skills in Python (or similar) with good practices: testing, version control, code review, refactoring.
- Hands-on experience with building production-ready agentic systems.
- Expertise in pushing the frontier of agent / LLM performance with techniques such as GraphRAG, memory systems and self-verification.
- Nice-to-have: experience building games in Unity or other engines is great, but not essential. I\'m sure you\'ll love games, too
The core stuff:
- Working from our East London office in-person 3 days a week.
- This is very much a founding team member role, so the equity level is going to be much more substantial than any other role. Equity can be further increased for top performers.
- We\'re not sugarcoating it; as a founding member, there will be a lot of responsibility and long hours. We\'re looking for a self-starting, extremely hungry individual who wants to build something incredible.
- We don\'t care about pedigree or titles; what we care about is what you\'ve managed to achieve above and beyond your peers.
Job Type: Full-time
Pay: £40,000.00-£80,000.00 per year
Work Location: In person
