AI Product Engineer

We started CoLoop in 2020 while still at university. We survived multiple cofounder breakups and all sorts of other near-death experiences. Despite that, we built our first product to $1M ARR, and then killed it after realizing the opportunity was too small and we didn’t have real PMF. We spent six months pivoting, and through that process came to understand just how important customer research is to finding a winning strategy. So we built a product to make it faster, cheaper, and better.

We also take seriously that you’re trusting us with your time. Three people who’ve worked at CoLoop have gone on to become YC founders. When you’re here, we’ll share everything we know and everything we learn along the way. All we ask is total commitment.

About the role: You’ll spend 90% of your time building. The rest is code review and talking to customers. At CoLoop, engineers own the full cycle. There are no product managers. You gather context, decide what to build, build it, launch it, measure the impact, and iterate. We track individual attribution in PostHog. The structure is flat. Influence is earned through skill and context. We’d rather you make a wrong call than wait to be told what to do. We run 5-6 week company-wide sprints with a decentralised approach and minimal meetings. We’re in our Oxford Circus office as often as possible (2 days min per week).

What we’re looking for:

• You’ve built and shipped AI products for real customers using TypeScript, Python, and the OpenAI/Claude/Gemini APIs.
• You’ve shipped complex (multi) agent systems to production.
• You have strong opinions about AI-augmented engineering. We use Claude Code, Codex, Conductor, Greptile, and more. We want someone who’s always finding the next thing that makes the team faster.
• You’re commercially minded. You think about what moves revenue, not just what’s interesting to build.
• You have an artisan’s mindset toward your craft. You can explain what makes Linear great, both as a company and product.
• You know when to cut corners and when to be rigorous. Move fast at the edges, build carefully at the core.
• You can explain what an AI agent is to both a 5 year old and the CTO of a customer.
• You want to get on call to fix a gnarly bug even though somebody else caused it.
• You recognise you have blind spots and want other people to point them out, and will happily return the favour.

Must-haves:

• 4+ years full-stack experience building web applications.
• 1+ years building generative AI applications.
• Fluent English.

Nice-to-haves:

• You’ve been a founding engineer at a high-growth company, or founded one yourself.
• You have significant experience mentoring engineers.
• You’ve built an AI side project with real traction.
• You’ve contracted or consulted on projects for enterprise-scale clients. People who’ve operated end-to-end for paying customers tend to think commercially, and that matters here.

What you get: You work directly with the founders and founding team on a daily basis. You join a tight-knit team of engineers: physics PhD, hackathon winner, and more. You build with cutting-edge generative AI, like our multi-agent system for qual content analysis. You become part of a community of top technologists in London. We regularly host dinners, hackathons, and CoLoop CoWorking from our office most Saturdays.

How to apply:

• 1-line introduction about yourself.
• 1-2 lines on what you’ve built and what you took away from it.
• A URL to a live project, repository, or open source contribution you made.
• 1-2 lines on your interest in CoLoop and how you’d contribute.

Interview process (~1-2 weeks): Screening interview (30 min), Technical interview (45 min), Presentation of work sample (30 min), In-person, paid contract (1 day), Decision.

If you’re excited about this role but don’t meet every requirement, we’d still love to hear from you. If you’d like to meet the team, we host CoLoop CoWorking from our office most Saturdays!