Join Sarvam, India's premier AI company, dedicated to building sovereign AI infrastructure and applications for India. As a Software Development Engineer on the Tejas team, you'll be instrumental in developing our context engine. This system is designed to empower AI agents with persistent memory and the ability to access and process an enterprise's diverse data, including documents, databases, and business systems. We are pushing the boundaries in knowledge and memory for AI agents, aiming for measurable improvements over traditional methods.

You will own significant aspects of our platform, working within a focused team environment. Your contributions will directly impact real customer deployments, providing rapid feedback on functionality. Key responsibilities include building and maintaining robust ingestion pipelines for large volumes of unstructured and structured data, ensuring reliability even with messy inputs or failing external services. You will also develop connectors to databases and enterprise systems, design data modeling strategies with ontologies and entity resolution, and create efficient retrieval tools with search, filtering, and citation capabilities. Furthermore, you'll build and refine agent memory systems at various organizational levels and iterate on quality improvements through rigorous evaluation.

This role requires 3 to 5 years of experience in building and managing backend or data systems in production, with a proven ability to troubleshoot and resolve issues. Proficiency in at least one backend programming language is essential; experience with Go is a significant advantage. Strong fundamentals in data modeling and PostgreSQL are necessary. Candidates should be adept at working with long-running pipelines, understanding retry and recovery mechanisms, and debugging search and retrieval issues. Familiarity with AI coding tools and a commitment to measuring and improving system quality through evals are also crucial.