From Blood Tests to Meal Plans: Building a Self-Correcting Health Agent with LangGraph

A developer built a self-correcting health agent using LangGraph, LangChain, and OpenAI that monitors laboratory biomarkers like cholesterol and uric acid, then dynamically rewrites lifestyle plans via OpenAI Function Calling. The agent uses SQLite for long-term memory and employs conditional logic to trigger a "Protocol Pivot" when abnormal lab values are detected, looping through analysis and plan revision rather than following a static linear chain.

Ever felt like your fitness app is just a fancy spreadsheet? You log a high uric acid result from your latest blood test, yet it still suggests a high-protein steak dinner for "gains." In the world of AI Agents , we are moving past static prompts. Today, we’re building a Self-Correcting Health Agent using LangGraph , LangChain , and OpenAI . This agent doesn't just chat; it monitors laboratory biomarkers like cholesterol and uric acid, maintains a long-term memory via SQLite , and dynamically rewrites your lifestyle plan using advanced OpenAI Function Calling . If you've been looking to master autonomous health agents and complex state management, you're in the right place. Let's dive into the future of personalized wellness. Unlike a standard linear chain, a health agent needs to "loop" and "reason." If the agent detects an abnormal lab value, it must trigger a specific logic branch to revise existing plans. Here is how the data flows through our LangGraph system: php graph TD A User Input/Lab Report -- B{Analyze Biomarkers} B -- Abnormal Found -- C Tool: Plan Rewriter B -- All Normal -- D Tool: Maintenance Plan C -- E Update SQLite Memory D -- E Update SQLite Memory E -- F Output Final Recommendation F -- G Wait for Next Input G -- New Data -- B To follow along, you'll need: In LangGraph, the State is the source of truth. We need to track the user's current health metrics and their active diet plan. python from typing import Annotated, TypedDict, List from langgraph.graph import StateGraph, END import operator class HealthState TypedDict : We use operator.add to keep a history of logs logs: Annotated List str , operator.add biomarkers: dict current diet plan: str revision required: bool The magic happens when the LLM decides whether your lab results require a "Protocol Pivot." We use OpenAI's Function Calling to detect high uric acid or cholesterol. python from langchain openai import ChatOpenAI from langchain core.messages import HumanMessage llm = ChatOpenAI model="gpt-4o", temperature=0 def analyze labs state: HealthState : biomarkers = state 'biomarkers' Logic to determine if intervention is needed needs fix = False if biomarkers.get "uric acid", 0 420: Example threshold needs fix = True return {"revision required": needs fix, "logs": "Analyzed lab results..." } def rewrite diet plan state: HealthState : prompt = f"Adjust the following diet plan based on these biomarkers: {state 'biomarkers' }. Current Plan: {state 'current diet plan' }" The LLM generates a new plan avoiding high-purine/high-fat foods new plan = llm.invoke HumanMessage content=prompt return {"current diet plan": new plan.content, "logs": "Diet plan updated for health safety." } Now, we connect the nodes. We use a conditional edge to decide whether to go to the rewrite diet plan node or straight to the end. workflow = StateGraph HealthState workflow.add node "analyzer", analyze labs workflow.add node "rewriter", rewrite diet plan workflow.set entry point "analyzer" Conditional Logic: If revision required is True, go to rewriter workflow.add conditional edges "analyzer", lambda x: "rewrite" if x "revision required" else "end", { "rewrite": "rewriter", "end": END } workflow.add edge "rewriter", END Compile with persistence SQLite from langgraph.checkpoint.sqlite import SqliteSaver memory = SqliteSaver.from conn string ":memory:" app = workflow.compile checkpointer=memory While this tutorial covers the core logic of stateful agents, building medical-grade or production-ready health platforms requires deeper security and more robust validation patterns. For advanced architectural patterns on deploying AI agents in high-stakes environments, I highly recommend checking out the WellAlly Blog . They provide excellent deep dives into integrating AI with real-world health data and "Agentic Workflows" that go far beyond basic tutorials. Let's test it with a scenario: A user with a "Keto" plan suddenly uploads a lab report showing high cholesterol. config = {"configurable": {"thread id": "user 123"}} initial input = { "biomarkers": {"cholesterol": 280, "uric acid": 450}, "current diet plan": "High protein, high fat Keto diet.", "logs": } for event in app.stream initial input, config : for value in event.values : print f"Update: {value.get 'logs', '' }" if 'current diet plan' in value: print f"New Plan: {value 'current diet plan' }" thread id , the agent remembers the previous state. analyzer node before the LLM even sees the data.Building autonomous health agents with LangGraph transforms a simple chatbot into a proactive health companion. By moving the logic into a directed graph, we gain control over the "reasoning loops" that LLMs often struggle with. What's next for your agent? Happy coding, and stay healthy 🚀💻🥑 Have questions about LangGraph or Agent memory? Drop a comment below