← back
Case Study + Deep Dive: Telemedicine Support Agents with LangGraph/MCP - Dan Mason
Takeaway
For regulated healthcare agents, pair LangGraph's visual orchestration with hybrid human-in-the-loop and MCP tools so doctors stay in control.
Summary
- Stride built an SMS-based telemedicine support agent for Aila Science (early pregnancy loss treatment) — replaces deterministic business logic with an LLM core that's more flexible while keeping doctors in the loop.
- Stack: Python + LangGraph (chosen because it's easy to visualize and explain to clients), Claude for the LLM, Node/React/MongoDB/Twilio frontend, AWS multi-region for EU HIPAA-style data residency.
- Hybrid system — humans in the loop for sensitive clinical decisions; LangSmith for tracing, custom eval harness built for the workflow.
- Code is mostly portable across Claude/Gemini/OpenAI via tool-calling + MCP; speaker explicitly invites criticism of design choices.
langgraphmcphealthcare
Original description
We've all seen website chat bots which can look up an order or answer a basic question -- but what does it take to build autonomous agents which manage long, delicate processes like multi-day medical treatments? In this workshop, we'll explore a workflow Stride built in partnership with Avila (https://avilascience.com/) that helps patients self-administer medication regimens at home. The stack includes LangGraph/LangSmith, Claude, MCP, Node.js, React, MongoDB, and Twilio, and rests on a foundation of treatment "blueprints" which LLM-powered agents use to guide patients to good outcomes. You'll learn how to: -Build a hybrid system of code and prompts that leverages LLM decisioning to drive a web application, message queue and database -Design and maintain flexible agentic workflow blueprints, with no special tools (just Google Docs!) -Create an agent evaluation system, which uses LLM-as-a-judge to evaluate the complexity of each interaction and escalate to human support when needed We'll also talk about the prompt engineered guidelines and guardrails which helps agents adhere to protocol as much as possible, while gracefully handling curveballs from the patient. Please bring questions -- we look forward to sharing our learnings on how to make agentic systems like this work in the real world! About Dan Mason Dan is a product and technology leader with unusually broad experience -- in 20+ years at companies like ESPN, Shutterstock, Viacom, NBCUniversal and a variety of startups and scaleups, he’s accumulated a wealth of knowledge about how digital product development works (and doesn’t), and is excited to apply those insights to reimagining teams and products in the age of LLMs. He is an engineer turned product manager with strong technical skills, and the teams he leads are highly cross-functional -- often including product, technology, design, PMO and data science. Dan leads Stride’s AI/LLM practice and is focused on thought leadership, code generation, workflow automation, and shaping and leading generative AI client engagements. He is also an active product coach and consultant, and a member of Docker’s Technical Advisory Group. Dan lives in New Jersey with his wife and three busy teenagers, and holds a BA in Computer Science and English Literature from Williams College. Recorded at the AI Engineer World's Fair in San Francisco. Stay up to date on our upcoming events and content by joining our newsletter here: https://www.ai.engineer/newsletter