AI Engineer (Full-Stack) — help build the taste layer for AI
Are you an engineer who cares about how products really feel? Have you ever shipped something and thought "yeah, that's not quite right" and then kept tweaking until it was? Do you have strong opinions about subjective quality, and the technical chops to encode them into systems?
A well-funded early-stage AI company is hiring 2-3 AI Engineers to help build something genuinely new - the systems that teach AI models what good looks like in subjective domains like design and creative.
They've raised a strong seed round from top-tier investors, they're a small team, and their customers include some of the most recognisable names in AI today. The work is backend-heavy with full-stack execution. So agents, evals, embeddings, crawling, synthetic data, plus the product surfaces where users actually interact.
What you'll do
▸ Build the AI core: agent harnesses, memory loops, self-improvement systems, evaluation pipelines, synthetic data generation
▸ Design retrieval at scale. Embedding and indexing infrastructure that handles millions of requests, plus the crawling and scraping systems that feed it visual data from across the web.
▸ Ship the product surface. Internal tools for data ops and expert annotators, external tools for clients, gamified flows like taste quizzes, leaderboards, and reward systems.
▸ Own the inference stack. APIs, serving, the tooling that makes everything reliable and fast under load.
▸ Invent, don't optimise. They're building infrastructure for something that doesn't exist yet, and they need someone who finds that energising.
The team
Small, exceptional, taste-obsessed. Team members leading start-up and hyper-scale product companies and backed by serious AI talent and on a mission to end AI slop and put more taste into the world.
The role is in-office in San Francisco, 4-5 days a week. They want builders who are bought into the mission and want to be in the room with the rest of the team while it gets figured out. This is a critical role and they're moving fast to fill it.
What they're looking for
✓ 4+ years of full-stack or backend engineering, with at least one tour at a sub-50-person startup or hyper-growth company
✓ You've shipped real AI systems - agent frameworks, RAG pipelines, embedding infrastructure, fine-tuning workflows. Whether at your job, in open source, or in serious side projects you can point to
✓ Strong Python, comfortable with React, cloud infra, and CI/CD
✓ Comfortable with crawling, scraping, and large-scale data processing
✓ Genuine curiosity about taste, design, and subjective quality - this is the part of the job that can't be faked
✓ Startup DNA. You operate well in ambiguity, ship fast, and don't need someone to hand you the spec
What they offer
💰 $180K–$300K + 0.25–0.75% equity
- A genuine stake in a company solving one of the most interesting problems in AI
- Visa sponsorship available for O1, STEM OPT, and H1B (if exempt from the $100k fee)
- Direct access to founders and real influence on the product from day one
- Work alongside engineers from Exa, Nous, Snorkel, Scale, and Palantir
- The chance to build infrastructure that doesn't exist yet, in a category being defined right now
