Sarvam is building the foundational sovereign AI platform for India, focusing on research, models, infrastructure, and applications to make AI genuinely work for the nation. Partnering with leading Indian enterprises and backed by prominent venture capitalists, Sarvam is at the forefront of AI innovation with significant population-scale impact.

This role offers the chance to design and implement critical system infrastructure, including MCP servers, on-prem deployment tooling, and data ingestion pipelines. You will be shaping architecture decisions that influence all future deployments, delivering real systems rather than prototypes, and taking ownership of their performance in operational environments.

Key responsibilities include designing and constructing backend systems for Sarvam's core AI components, such as MCP servers and document ingestion pipelines. You will develop robust API layers connecting the AI stack to frontends and operational systems. Implementing data ingestion for unstructured formats like PDFs and audio transcripts is crucial. The role also involves building for secure, constrained deployment environments, writing well-tested code, and collaborating with cross-functional teams. Supporting strategic deployment engineers with troubleshooting and operational tooling is also part of the scope.

We are seeking backend engineers with 3-6 years of experience in production systems. Proficiency in Python, particularly FastAPI, asyncio, and Pydantic, is essential, with Go or Rust being a significant advantage. Strong system design skills, including distributed systems, job queues, and async processing, are required. Experience with containerization technologies like Docker and Kubernetes, along with database knowledge (PostgreSQL, Redis, vector databases), is necessary. Familiarity with LLM APIs and building agentic pipelines using frameworks like LangChain is also expected.