About the role
At Sarvam AI, a leading full-stack AI company based in Bengaluru, the engineering team focuses on building advanced AI infrastructure and applications tailored for the Indian market. The role is part of a small, agile team responsible for developing Sarvam's context engine, which enables AI agents to effectively access and remember enterprise data across various sources. This platform supports real customer deployments, allowing engineers to see the impact of their work quickly and collaborate closely with teams working on foundational models and document AI.
The Software Development Engineer will primarily build and maintain the ingestion pipeline that processes large volumes of enterprise documents and data, ensuring reliability despite messy inputs or external failures. They will design data models and ontologies to unify information from multiple sources, develop fast and cost-effective retrieval tools for AI agents, and implement memory systems that track information at user, team, and organizational levels. The role involves continuous evaluation and iteration to improve answer quality, as well as collaboration with product and deployment teams to generalize customer-specific solutions.
Candidates should have 3 to 5 years of experience building and operating backend or data systems, with strong skills in at least one backend language, preferably Go. Proficiency in data modeling, PostgreSQL, long-running pipeline management, and search or retrieval debugging is essential. Familiarity with AI coding tools and a focus on quality measurement are important. Additional experience with knowledge graphs, memory systems for AI, document processing, enterprise system connectors, or startup environments is advantageous. The position is onsite in Bengaluru and involves working with technologies like Go, Temporal, OpenSearch, and Postgres.