About the client
A confidential client in technology. The company is named to you once you and the recruiter have spoken.
The role
Build multi-agent AI systems in Hyderabad for a technology company. You will design agent workflows with LangGraph, shape prompting and retrieval strategies, and benchmark leading foundation models. The role suits a senior engineer who has taken LLM applications into production.
The team builds AI services in which several agents pass work to each other to complete a task. You will architect these orchestration patterns, using LangGraph to design skills-based workflows that run reliably at scale.
A typical day includes design reviews with colleagues and work on prompting and retrieval strategies that give grounded, dependable outputs. You will compare foundation models for accuracy, latency and cost. You will also mentor junior engineers in the Hyderabad team on prompt engineering, agent design patterns and production-quality Python.
You will work with architects and product owners based in Europe, so comfort with cross-timezone work and leading technical discussions matters. Compliance with the EU AI Act, DORA and internal DevSecOps standards is part of the engineering bar.
What you bring: a degree in computer science, AI or a related field; expert Python; hands-on work with LangGraph and LangChain; RAG and agentic workflow design; working knowledge of Claude, GPT and Gemini, including fine-tuning, context window management and model evaluation; and at least 2 years of LLM applications in production. Experience with Model Context Protocol, GitHub Actions, Terraform, Kubernetes, regulated financial environments or German is a plus.
The role is full-time, in the office in Hyderabad, on day shift. It needs 5 years or more of professional AI or ML engineering.
You
Must have
Nice to have
On offer
- Full-time role
- Day shift
- Notice period ≤ 30 days
Summary
- Location
- Hyderabad
- Work mode
- Office
- Experience
- 5–10 yrs
- Posted
- Today
What happens after you apply
The recruiter working this role reads your application. If it fits, they call you before anything is shared with the company. You see every stage in your account, in plain words.