Founding Engineer, Agent Systems

We're building governance and security infrastructure for a world of increasingly capable and widely proliferated agents.

The world is undergoing a fundamental shift: every day more work is being done autonomously by AI. This is true for enterprises but also for the adversarial actors who seek to disrupt their operations through cyber attacks. The only way to keep up is to use AI intelligently to govern internal deployments and defend against offensive threats. Helm Guard will be the platform which businesses use to do both.

We've grown to seven-figure revenue within 9 months of product launch, on the back of multi-year contracts with leading enterprises in financial services, regulated technology, and healthcare. Our founders come from Palantir and academic institutions:
Oxford, Stanford, and ETH. We're backed by leading UK and US institutional investors and exceptional angels from Meta, Isomorphic Labs, Palantir, SpaceXAI, and more.

We're hiring across founding-team roles for people who want outsize impact, the influence over direction and culture that comes only from joining this early, and pre-Series A equity upside.

• In-office 4 days a week (with flexibility for things like school drop-off)

• Wanting to own outcomes and to take on the hours required to do so

• Having an existing right to work in the UK (we do provide visa support for applicants already working in the country)

We already have one of the most robust and high-performance agent scaffolds in our vertical. You’ll help make it one of the best AI agent platforms being shipped today. In addition, you’ll help define the primitives that fast-growing startups and massive government organisations use to govern agents.

You own the backend for our agent platform. This means you are responsible for the cost, performance, and reliability of the agents we use. What’s involved in that can expand with your ambition and our progress as a company. On day one, it involves further improvements to our evals, continued integration of our AI security primitives, and iteration on our harness.

In the future, it will also likely involve research under our CTO to further improve our agent delivery and help govern our customers’ agents.

Importantly, this role is mainly engineering, with some room for focused research. We expect candidates to have run agents in production and to be ready to build for production (not research) workloads.

• Agent scaffolding: tool use, context management, sandboxing, prompt-injection defence

• Evals for security and GRC workflows

• Reliability infrastructure: retries, fallbacks, circuit breakers, prompt versioning

• Experience with backend engineering in Type Script or a comparable language, with 1–2 solid projects involving production LLM features

• Experience with agent frameworks, tool calling, and multi-step orchestration

• Good taste in eval construction and grading

• Strong systems thinking: async, queues, idempotency

• Comfortable reading and implementing research from AI venues like ICML, NeurIPS, TMLR or ICLR

• Background in GRC or security

• Background in AI security, safety or governance

• You haven’t deployed an AI agent consumed by someone other than yourself and iterated on it using evaluations as a scorecard

• You wouldn’t be able to confidently answer relatively niche questions about Anthropic’s API

• You aren’t able to name an inference provider outside of the major labs

• You've never had to make an agent system cheaper

• You’ve shipped an LLM feature but not built an eval for it

We value a diversity of perspectives and experiences. We also hold a small set of core beliefs that reflect how we operate, and…