Full-Stack AI Developer

About the role

You will build the whole product, not a slice of it: the interface, the API, the data layer and the AI services that sit underneath. Our work is AI that has to survive real users, so you will care as much about the rollback path and the p95 latency as about the feature itself. You will own features from a first scope through to production and the on-call that comes with it.

What you will do

• Build and ship full-stack features for AI products, with a React or Next front end and a Python or Node service behind it.
• Integrate LLMs, retrieval pipelines and third-party APIs into clean, observable services that hold up under load.
• Model the data, design the schema and wire the vector and relational stores the AI reads from.
• Instrument what you ship: logging, tracing and the dashboards that tell you when something is drifting in production.
• Own the deployment path, including CI, environment config, rollbacks and the cost and latency budgets for your service.
• Review code, write the tests that matter, and keep the shared stack small and consistent across pods.

What we are looking for

• 3 to 6 years building and shipping web applications, with real production ownership rather than only feature tickets.
• Strong JavaScript or TypeScript with React or Next, and a backend language you are fluent in, ideally Python or Node.
• Solid grasp of relational databases and API design, plus the basics of containers and one major cloud.
• Hands-on experience wiring an LLM or external API into a product, including the error handling that keeps it standing.
• A production mindset: you reach for the eval, the rollback and the metric before you reach for the demo.

Nice to have

• Experience with FastAPI, pgvector or a managed vector database, and a queue or streaming system.
• Familiarity with RAG patterns and agent frameworks such as LangGraph.
• Comfort with infrastructure as code and observability tooling.

What we offer

• Whole-product ownership, so you ship features instead of fragments.
• A modern, deliberately small stack shared across pods, with senior engineers to learn from.
• AI work that reaches production, measured against the five numbers in the Production-First method.
• A hybrid seat at our Noida delivery center, with a team that has built in-house since 2017.