AI Builder - Full Stack - Groupon

Groupon connects 42 million customers with local experiences — restaurants, events, wellness, and travel — and the million-plus merchants who deliver them. Our mission is to get people offline and into the real world at prices that make it possible. We're an AI-native company in the middle of a platform transformation, moving from a deals marketplace to an experience discovery platform that works for customers and merchants at the same time.

We are now looking for AI Builder - Full Stack to join our team.

Groupon's internal teams are building AI agents to automate operations, but every team is solving the same infrastructure problems in isolation — no shared platform, no shared patterns. This role builds the foundation: a proprietary agent creation layer that lets engineers across the organisation ship autonomous AI solutions without starting from zero. When this hire does their job well, agent development time across Groupon drops from weeks to days and the company moves faster on every initiative that depends on AI.

Who you'll work with

Report to Daniel Lledó. A team of [2] engineers working on foundational AI infrastructure. This team ships experimental platform code into production fast — if you need fully scoped requirements before starting, this environment will frustrate you.

What you'll be doing

• Building and expanding the proprietary agent platform so that non-ML engineers across Groupon can ship AI solutions without reinventing core infrastructure
• Engineering reasoning flows and integration layers that connect agent systems to production services, with your name on the reliability of every downstream consumer
• Designing failure-handling and redundancy patterns so agents degrade gracefully rather than failing silently under real conditions
• Building the tools — intent retrieval, context management, API routing — that every future agent in the company depends on
• Owning code quality in a fast-moving codebase, catching the patterns that create future incidents before they reach the teams building on top of you
• Embedding AI-native development tooling (Claude Code, Cursor) into daily workflows as standard practice, not an experiment

What success looks like

• Day 30: Environment set up, first production bug fixed, first PR merged into the agent platform. You understand the existing architecture well enough to explain its tradeoffs to a new hire.
• Day 60: First independent feature shipped — an agent building block that at least one other team is already using. You've identified one structural risk in the platform and proposed a fix.
• Day 90: Operating without supervision. The platform is handling measurably more agent deployments than when you joined, and you own at least one cross-cutting concern end-to-end.

What this role demands

• Production TypeScript and Node.js across backend and frontend (Express, React, Next.js, SQL) — shown through shipped software, not side projects
• Hands-on experience with Chain of Thought reasoning, agent architectures, and RAG — not theoretical familiarity
• Docker, Kubernetes, and Linux in production — you've debugged real incidents, not just run tutorials
• AI-first working practice: Claude Code or Cursor is already part of how you work, not something you're open to trying
• Ability to ship into a shared platform used by multiple teams — you think about downstream impact before merging

How we operate

• Extreme Ownership: The platform you build becomes the dependency every AI team relies on. Your responsibility doesn't end at the PR.
• Speed Over Comfort: The agent space is moving fast. Working drafts shipped and iterated beat polished specs that arrive late.
• Impact Obsessed: The measure of success is how many teams are shipping AI solutions faster because of the infrastructure you built.

#LI-KH1  #LI-Hybrid #marci