Location: Hyderabad (on-site) | Experience: 3–5+ years | Employment type: Full-time
Edvak builds software for healthcare practices in the United States. We are now bringing AI agents into our product, and we are looking for an engineer to lead that work.
As our Agentic AI Engineer, you will own the AI side of Edvak: designing, building, evaluating and running agents in production. You will work closely with our software engineers and domain experts, who will support you in building and shipping.
Build LLM agents that use tools and APIs to complete multi-step tasks
Design prompts, context and structured outputs that are accurate and cost-efficient
Establish evaluations and monitoring so that changes in quality are caught early
Build in safety by design: untrusted inputs, permissions, audit logging and sensitive data handling
Guide engineers on AI design and review agent code
3–5+ years of software engineering experience, with strong Node.js and TypeScript
Hands-on experience building LLM applications or agents used by real users
Practical experience with tool calling (function calling), structured outputs and multi-step agent loops
Working knowledge of the Claude or OpenAI APIs, including cost, latency and rate limits
Experience with evaluations and tracing tools (Langfuse, LangSmith, promptfoo or similar)
Understanding of prompt injection and the safe handling of sensitive data
Agentic AI is a young field. We value what you have built over how many years you have spent doing it.
Familiarity with frameworks such as LangChain, LangGraph or the OpenAI/Anthropic Agent SDKs, and with RAG or vector search
Experience with MCP or sandboxed code execution
Familiarity with MongoDB or NestJS
Experience in healthcare or with HIPAA
Python for evaluation and data work
Competitive compensation
Real ownership of AI at Edvak
Work on problems that matter in healthcare
Edvak is an equal opportunity employer. We welcome applicants of all backgrounds.