AI Research Engineer

Train Indic LLMs from scratch. Strong DL + distributed training background.

As a AI Research Engineer at Sarvam AI, you'll join a small, senior team building AI-native saas products from the ground up. You'll work daily with Python, Claude to design, ship, and scale features that customers use every day. This is a high-ownership role — you'll have direct input on what we build and how we build it.

• Architect end-to-end saas workflows that integrate LLMs (OpenAI, Anthropic, Gemini) with internal systems
• Optimise inference cost, latency, and reliability across services
• Collaborate with product, design, and customer teams to translate business problems into AI solutions
• Own the full lifecycle of features — from spec, prototype, A/B test, to rollout
• Lead architectural reviews for new SaaS initiatives
• Instrument observability — eval pipelines, prompt versioning, cost dashboards
• Write clean, well-tested code in Python / TypeScript and review PRs from peers

• Proficiency in Python, TypeScript, or both
• Strong hands-on experience with Python / Claude in real production environments
• Bias for shipping — comfortable owning ambiguous problems end-to-end
• Experience with prompt engineering, evals, and prompt-versioning workflows
• Solid understanding of REST APIs, webhooks, queues, and event-driven design
• Bachelor's degree in CS, Engineering, or equivalent hands-on portfolio
• Comfort working with vector databases (pgvector, Pinecone, Weaviate)

• Past work in the SaaS domain
• Built and operated agentic systems in production
• Familiarity with Cloudflare Workers / serverless edge runtimes
• Open source contributions to AI / dev-tools projects

• Work on problems that genuinely move the needle in SaaS
• Direct ownership — your code reaches real users within days, not quarters
• Learn from a team that has shipped at scale and built with Python, Claude, Claude from day one
• Tiny team, big surface area — wear multiple hats and grow fast