Applied AI Engineer, Sarvam Agents

Ovii's Interpretation of the Role

The Applied AI Engineer will design, build, and operate end‑to‑end AI agents that automate workplace workflows. You will own the full production stack—from prompt engineering and tool integration to reliability, scaling, and observability—delivering AI features to real users.

Role Snapshot

• End‑to‑end AI agent development
• Production reliability ownership
• Prompt & tool integration design
• MCP server scaling
• Memory & context engineering
• Evaluation & monitoring pipelines

Must-Have Requirements

• Python backend development
• LLM API fluency (prompt design, function calling)
• Agentic framework experience (LangGraph, ADK)
• Production experience with MCP servers
• RAG techniques (chunking, embeddings, vector DBs)
• Building and running AI evals
• Cost and latency optimization for LLM workloads
• Observability tools (Datadog, Signoz)
• OAuth & SaaS API integration
• PostgreSQL or MySQL
• Redis
• backend development
• LLM integration
• production AI feature delivery

Nice-to-Have Signals

• Open‑source contributions in LLM/agent ecosystem
• Familiarity with eval frameworks (LangSmith, Braintrust)
• Human‑in‑the‑loop / approval flow design
• Early‑stage startup experience
• open‑source contributions
• early‑stage startup experience

Work Setup

• Location: Bengaluru, India
• Work mode: ONSITE
• Employment type: Full-Time

Not Specified in JD

• Visa sponsorship
• Salary range
• Remote eligibility
• Education requirement
• Certifications
• Relocation
• Notice period
• Travel
• Security clearance
• Coding test
• Portfolio
• GitHub
• Writing sample
• Cover letter

What You'll Likely Work On

• Design and implement full agent flows, prompting, tool schemas, memory handling and deployment pipelines
• Maintain agent reliability in production, including debugging, guardrails and incident response
• Build the runtime layer: state management, retries, scheduled triggers and long‑running execution
• Create and manage OAuth‑based connectors to third‑party SaaS services
• Scale MCP server infrastructure that powers agent tool access
• Engineer memory systems, balancing working, long‑term and retrieval memory
• Develop multi‑tenant, permissioned agent instances with audit trails and admin UI
• Construct evaluation test harnesses and A/B testing workflows for prompt iteration

Good Fit If You Have

• Open‑source contributions to LLM or agent ecosystems
• Experience in early‑stage or growth‑stage startups
• Familiarity with eval platforms such as LangSmith or Braintrust
• Designing human‑in‑the‑loop approval flows for AI systems