Ashish Jain

Ashish Jain is a research engineer at Sarvam AI, where he works on post-training and capability development of large language models tailored for enterprise use cases. His work focuses on building reliable, high-quality conversational AI systems by combining model improvements with robust real-world deployment infrastructure.

At Sarvam AI, he has led the development of conversational and agentic capabilities on top of foundational LLMs, including designing large-scale data pipelines for multilingual, multi-turn enterprise datasets grounded in real-world interactions and tool use. He has also built memory layers for conversational agents to enable better continuity and context retention across long-running workflows.

Before this, Ashish worked as an applied AI engineer, leading AI deployments across BFSI, digital-native companies, and government organisations.

One of his key milestones includes leading the deployment of an AI solution for one of India’s largest banks within a three-month timeline, enabling customers—even in tier-3 towns—to interact in their native languages and receive accurate query resolution.