About Eka Care

Eka Care is one of India's leading HealthTech startups in Agentic AI, spanning ambient clinical documentation (Eka Scribe) all the way to an EMR trusted by thousands of doctors across clinics and hospitals. We sit at the intersection of AI and the doctor's daily workflow, and our systems are already live at scale.

You'll be joining the core team building the next layer of that intelligence:  Agentic flows for triage, appointment booking and clinical routing. Real problems. Real doctors. Real impact.

Areas you will learn about & contribute

1. Building the agentic loop for slot discovery, preference matching, and confirmation, integrating with Eka's scheduling APIs via MCP tool definitions.
2. Setting up of prompt management system : version control for prompts, eval datasets and regression testing across model updates.
3. Design the graceful degradation layer : when does the agent escalate to a human? What context does it hand off, and in what format?
4. Define and implement MCP tool specs for Eka's internal services (patient lookup, slot availability, medication data) , making them consumable by any LLM agent layer we build on top.
5. Run a structured comparative analysis of existing health bots across conversation quality, triage accuracy, drop-off points and feed findings back into product decisions.

Who This Is For

•     You've built at least one agent or RAG system &  not just followed a tutorial

•     You care about prompt quality the way engineers care about code quality

•     You've lost sleep over why a model gave the wrong output and actually debugged it

•     You want to work on something that ends up in front of real doctors and patients in India

Include your GitHub profile and links to any live demos or projects. We read every application & do not use any  Automated Screening Filters. If your work speaks, we'll reach out within a week.