Agent Engineer

Overview

Build production AI agent systems on the Sarvam stack, covering design-time scenario mapping, quality measurement, integrations, and reliable scaling in live environments.

What you'll do

• Build production agents including workflows, tool integrations, evals, and system design.
• Map the scenarios an agent must handle, including edge cases, by design.
• Build fast proofs of concept for customers before scaling.
• Integrate agents into systems like core banking, CRM, ticketing, and telephony.
• Own quality end to end by defining success criteria and running evals to close gaps.
• Version and regression-test agent artifacts to keep quality stable.
• Tune agents for accuracy, cost, reliability, and performance under real load.

What you'll need

• 3+ years of experience building something real users depended on.
• Experience shipping AI to production and owning it after launch.
• Serious experience with LLMs and agents, including failure modes at scale.
• Use agentic frameworks and open/closed source models.
• Engineering discipline with versions, tests, regressions, and reproducibility.
• Ability to define eval metrics and build pipelines to separate signal from noise.
• Strong Python and broad stack range across backend, data pipelines, APIs, and cloud infrastructure.

Details

• Location: headquartered in Bengaluru.