Senior Engineer - Backend Typescript

The systems that power how merchants create, price, and submit deals are being rebuilt from scratch. [X deals processed daily / X% of GMV] flows through infrastructure built for a different era. This role owns the backend engineering of the replacement: TypeScript-native, observable end-to-end, with AI tooling built into the development cycle from the start. When this hire is effective, the commercial team stops being blocked by the platform and merchants can move in hours, not days.

Who you'll work with

You'll report to Mayank Jain, Manager, Software Engineering II, who owns full end-to-end tool focused on developing  the backend components for deal creation in Groupon IQ. The team is 4 engineers operating across Europe and Bangalore, India. Engineers make and defend their own technical calls here. If you need a settled roadmap and clearly scoped tickets to feel productive, this environment will frustrate you.

What you'll be doing

• Migrating core deal management services from Java to TypeScript/Node.js, reducing operational complexity and giving the team a codebase it can move quickly in
• Designing and building the APIs that commercial operations and merchant-facing products depend on, with written trade-offs and clear ownership at every layer
• Owning SLOs and observability for your systems: writing the dashboards, setting the thresholds, and being the person who owns the fix when something degrades
• Shipping features end-to-end: design, implementation, rollout, and post-production monitoring, without handoffs at each stage
• Using Cursor and Claude as working infrastructure — not side experiments: concrete decisions about which tasks you delegate to AI tooling and how that changes your output
• Writing RFCs and design docs for architectural decisions, so the team has a written record of trade-offs before committing to an approach
• Setting the quality bar for your area through code reviews, testing strategy, and documentation that means someone else can own this system in two years

What success looks like

Day 30: You've shipped at least one change to a production service, you understand the data model underpinning deal creation and submission, and you've named one observability gap you want to close.

Day 60: You've led a design review for a non-trivial feature, you're regularly using Cursor and Claude in your development cycle and can describe concretely what changed because of it, and your work doesn't need a check-in to move forward.

Day 90: You've completed or meaningfully progressed a Java-to-TypeScript migration for at least one service, you're the named owner of the systems, and you've written a design doc that influenced a decision beyond your own sprint.

What this role demands

• Ship production-quality TypeScript and Node.js at scale: code that survives high traffic, passes review, and doesn't require its author to be the only person who can debug it
• Read and refactor Java safely enough to migrate legacy services without introducing regressions — you don't need to be a specialist, but you need to be honest about what you're changing
• Trace a latency regression across distributed service boundaries, write the root cause analysis, and implement the change that prevents recurrence
• Write an RFC or design doc that a non-author engineer can read, challenge, and act on without needing you to explain it first
• Use AI coding tools as a genuine productivity multiplier: you've made specific choices about what you delegate to Cursor and Claude, and you can show the difference in your output

How we operate

Extreme Ownership: Your name is on the system. If an SLO degrades or a migration goes sideways, you own the diagnosis and the fix. Nobody waits to assign it to you.

Speed Over Comfort: The TypeScript migration is happening. You'll ship before everything is perfect and make calls with incomplete information. Engineers who wait for full clarity slow the team down.

Disciplined: Observability and documentation aren't optional. The pace of transformation makes rigor more important, not less. A system with no dashboards and no named owner is a production incident waiting to happen.

#LI-Hybrid #marci