Agentic AI for Clinical Coding: From Clinical Notes to Medical Codes

Healthcare organizations generate large volumes of unstructured clinical documentation that must be translated into standardized medical codes for billing, reimbursement, reporting, and analytics. Traditional coding workflows can be time-consuming and often require significant manual review. In this presentation, we will explore how Agentic AI and multi-agent architectures can assist clinical coding by coordinating specialized AI agents, each responsible for a specific part of the coding workflow. We will demonstrate an end-to-end Agentic AI workflow: Clinical Note → Clinical Information Extraction → Semantic Code Retrieval → Medical Code Selection → Validation → Human Review. A practical Python demonstration will show how technologies such as LangGraph, Large Language Models (LLMs), OpenAI embeddings, FAISS vector search, and SQLite can work together to create an intelligent clinical coding workflow. Rather than allowing an LLM to simply generate a medical code, the architecture uses retrieval, specialized agents, validation, and human oversight to create a more controlled, explainable, and reliable AI-assisted workflow. For more information look at the series of biotech papers at “Artificial Intelligence Applications in Biostatistics, Bioinformatics and Computational Biology by Ernest Bonat, Ph.D.”

Topics include
The presentation will introduce a multi-agent architecture consisting of specialized agents for:

1. Clinical Information Extraction – Extract diagnoses, symptoms, procedures, anatomy, and other relevant clinical information from unstructured notes.
2. Semantic Retrieval – Use embeddings and vector search to identify medical code descriptors that are semantically related to the documented procedure or service.
3. Medical Code Selection – Evaluate retrieved candidates and identify the codes best supported by the clinical documentation.
4. Coding Validation – Independently verify that proposed codes are supported by the clinical note and retrieved coding information.
   5. Workflow Coordination – Orchestrate agents, tools, state, and validation using an Agentic AI framework.

Who should attend:
This presentation is designed for AI/ML engineers, data scientists, software developers, bioinformatics professionals, healthcare technology professionals, researchers, medical informaticists, and anyone interested in practical applications of Agentic AI in healthcare and life sciences. No advanced knowledge of Agentic AI is required. The presentation will introduce the architecture step by step and conclude with a practical multi-agent clinical coding demonstration.

Agenda:
· 5:30 – 6:00 pm: Networking and refreshments
· 6:00 – 6:10 pm: Welcome!
· 6:10 – 7:30 pm: Presentation and open discussion
· 7:30 – 8:00 pm: Networking

Location: Entrepreneur Collaborative Center, 2101 East Palm Avenue, Tampa, FL 33605

Parking: Free parking is available in the lot directly north of the ECC building. Please do not park immediately adjacent to the facility.
RSVP: Seating is limited — please RSVP early. If you're bringing a guest, have them RSVP separately so we can plan accordingly.

Speakers:
Ernest Bonat, Ph.D. — Senior GenAI Engineer specializing in Machine Learning systems and AI assistants for Biostatistics , Bioinformatics, Computational Biology, and Healthcare.