Senior Product Engineer (AI-Native)

About the Role

This is an engineering role first. You'll design schemas, build APIs, ship UI, and operate production systems on GCP, and you'll be judged on the quality of what you ship. On top of that foundation you'll own the product side of your work too: you'll take a product question ("why did this model stop converting?"), turn it into a spec, build it end to end, and watch the metric move. A recent example of the shape of the work: our market-prices system, where one engineer owns everything from the scraper data contract and the partitioned Postgres tables to the upload flow with dry-run validation and the calibration report that tells us when a rival outbids us.

The role sits deliberately between product, design, and engineering. There's no separate PM writing your tickets: you'll frame the problem, write the spec, and define what success looks like, then build it. Our design system lives in code, so you'll make the design calls and ship polished UI without waiting for a handoff. And you'll build for two kinds of users: people and AI agents. Our trade-in flow already runs inside Claude through our MCP connector, and a growing share of our operations run as agent workflows the team writes as skills.

A note on AI-native, because it cuts both ways: building with agents daily is non-negotiable here, but AI multiplies engineering skill, it doesn't replace it. You need the depth to review, correct, and own everything an agent writes, because you could have written it yourself. Vibe coding doesn't survive contact with a production pricing system.

What You'll Do

• Own features end to end: spec, Postgres schema, NestJS and Cloud Functions APIs, Next.js UI, deployment, and the metric afterwards.
• Build the systems behind pricing: market-data ingestion with dry-run validation, append-only observation datasets, and calibration reports ranked by competitor gap.
• Ship customer-facing product across the quote, checkout, and merchant-embedded flows, making the design calls inside our design system.
• Design and implement APIs and integrations for merchant partners (Shopify, Magento, any custom API integration).
• Do the product job for what you build: write the specs, prioritize with the CTO, define success metrics, and follow the results after launch.
• Extend our agent surface: write skills and plugins for the team, build MCP servers for internal tools, and improve the public connector that lets customers trade in devices inside Claude.
• Automate operations with agents: alert triage, changelogs, reports, data checks. If you do something twice, you teach an agent to do it.
• Write tested, production-quality code (Jest, Playwright) and hold a high bar in reviews.

Apply if you

• Have 5+ years of professional experience building and running production systems with TypeScript, Node.js, React/Next.js, and PostgreSQL. Real systems with real users, not demos or prototypes.
• Have built and shipped production APIs and integrations, ideally on serverless GCP (AWS also valuable).
• Care about data modeling: you understand why append-only tables, point-in-time correctness, and partitioning matter for data that will train models later.
• Work AI-natively, and this is non-negotiable: agentic coding tools (Claude Code or similar) are your daily driver, and you know the ecosystem well enough to write your own skills, plugins, and MCP servers rather than just chat with a model.
• Can defend every line an agent writes for you: you review it, correct it, and own it in production.
• Have product sense: you can write a spec, scope an MVP, challenge a prioritization call, and own a metric.
• Have design sensibility: you sweat UI details and can ship a polished interface from a design system without a handoff.
• Are a strong communicator who works independently in a small team.
• Are fluent in English (written and spoken).

Even better if you have

• Machine learning experience. Our market-price dataset is being built append-only and point-in-time correct precisely so a payout model can be trained on it, and the engineer who owns pricing calibration will take it from deterministic gap math to a learned model.
• LLM product features in production: tool use, structured outputs, evals.
• Experience in e-commerce, fintech, or marketplaces.