Forward Deployed Software Engineer

Sarvam AI is building India's full-stack sovereign AI platform across speech, text, translation, and conversational agents. As a Forward Deployed Software Engineer (2-5 years), you deploy Sarvam's AI agents into client and partner production environments end-to-end.

Responsibilities:

• Build, deploy, monitor, and evaluate conversational AI agents across channels — Voice, WhatsApp, website.
• Own end-to-end deployment of Sarvam's dubbing and live translation APIs into client and partner production.
• Scope and design end-to-end technical solutions for client requirements, translating business needs into production-ready architectures.
• Build MCP (Model Context Protocol) servers and agentic backend infrastructure that power Sarvam's AI agents.
• Design and maintain data pipelines using messaging systems like Kafka, SQS, or RabbitMQ.

Requirements:

• Python fluency across the full stack.
• Deep understanding of how LLMs and agents work — context windows, tokens, embeddings, RAG, tool calling.
• Experience designing and implementing multi-agent architectures, including memory management.
• Ability to set up eval pipelines and define meaningful quality metrics for agentic AI systems.
• Familiarity with LangChain, LlamaIndex, CrewAI, LangGraph; Kafka/SQS/RabbitMQ; SQL/NoSQL; voice/speech models (TTS, STT, ASR).