Software Engineer – Building at the forefront of AI
UK/London – | Early stage AI Startup | High Impact, High Curiosity Role
If you’re a curious engineer fresh out of university (or with 1–3 years of experience) and obsessed with AI, you will not find a more exciting position or business than this! We’re hiring multiple positions and are open to a range of seniorities)
We’re partnering with an extremely well-funded, seed-stage AI company building something foundational: interactive, resettable digital environments where AI agents learn through experience, not just static datasets.
Think: instead of just training on text, AI learns by actively using software, from video editing tools to complex simulation platforms, inside high-fidelity digital sandboxes.
This role sits at the heart of that mission.
The Role
You’ll design the actual training curriculum for AI systems.
That means:
- Turning real-world software usage into hundreds of structured problems for AI models to solve
- Writing prompts, defining expected behaviours, and building verifiers
- Using Python + LLM APIs + clever tooling to generate tasks at scale
- Experimenting constantly to improve how models learn
It’s hands-on, experimental, and highly technical. There’s rarely a single “right” answer – progress comes from trying things, iterating fast, and finding leverage.
If you like building scrappy prototypes, hacking together clever solutions, and seeing immediate results, you’ll thrive here.
What You’ll Be Doing
- Creating structured problem sets that train AI models on real software
- Building Python-based verification systems to evaluate AI behaviour
- Collaborating with domain experts to translate messy real-world workflows into structured learning tasks
- Designing tooling, scripts and agentic workflows that multiply output over time
- Running fast feedback loops and QA to ensure quality before delivery
What we’re Looking For
- 1+ years using Python to solve real problems
- Software engineering fundamentals
- Experience using LLM APIs or working with agent-style systems
- A hacker mindset – pragmatic, experimental, not obsessed with over-engineering
- Comfort with shipping fast, learning fast, iterating fast
- Genuine intellectual curiosity – you enjoy diving into new technical domains
This would suit someone
- Fresh out of university with strong internship experience
- Or 1–2 years into their career and wanting serious responsibility
- Who wants exposure to frontier AI work without 1,000 layers of corporate process
Why This Is Special
- You’ll work on one of the most technically interesting problems in AI right now
- Small team, massive impact — your work ships directly to leading AI labs
- Seed-stage but well-funded and backed by top-tier founders and investors
- High-trust, low-bureaucracy culture
- Competitive salary + per-deliverable bonuses
This isn’t a “maintenance engineering” role. It’s experimental, foundational, and high-agency.
If you’re the kind of person who
- Builds side projects for fun
- Breaks tools to understand how they work
- Gets excited about LLMs and agent systems
- Wants to be early in something that could matter
#J-18808-Ljbffr…