AI Engineer Needed for Medical AI Agent / Clinical Text Parser (Python & LLMs)

We are seeking a talented AI/ML Engineer to develop a lightweight, functional Proof of Concept (POC) for a Medical AI Agent. The goal of this micro-project is to build a scoped Python script or agentic workflow that processes clinical text (such as anonymized patient notes or medical charts) and extracts structured medical insights. This is a well-defined, milestone-based project. Successful delivery of this POC will lead to a larger, higher-budget contract to build out the full platform. Key Responsibilities Structured Medical Extraction: Build an LLM-backed pipeline to parse unstructured clinical text and extract key entities, including diagnoses, prescribed medications, dosages, and active symptoms. Standardized Output: Ensure the agent maps extracted data into a clean, predictable JSON schema (e.g., matching semi-structured formats like synthetic FHIR-lite resources). Agentic Logic/Tooling: Implement basic validation or routing logic (e.g., flag potential drug-to-drug interactions using a mock lookup dictionary or external medical API). Clean Delivery: Provide modular, well-documented Python code that can be easily run and verified locally. Technical Requirements Language & Stack: Python 3.10+ using LangGraph, LangChain, or direct structured outputs via OpenAI/Anthropic APIs. Medical AI Awareness: Familiarity with parsing medical jargon and ensuring strict adherence to deterministic JSON structures (e.g., utilizing Pydantic for schema enforcement). Privacy Mindset: Understanding of data handling best practices (Note: Only synthetic, entirely anonymized data will be used for this POC). Timeline: 3–5 days from contract start. Screening Questions How do you ensure an LLM reliably outputs strict, validated JSON (like Pydantic schemas) when processing complex medical notes? Have you worked with medical text or health tech APIs before? Please briefly highlight a relevant project. Confirm your availability and budget.