✅ Inspiration

To empower individuals and families with easy access to local health, social, and wellness services through a conversational AI experience, especially those underserved or unaware of available community support.

✅ What it does

WellnessNavigator Agent enables users to ask natural language questions to get real-time insights on available wellness services, benefits, and local community programs. It summarizes patient benefit details and matches individuals to nearby resources using AI.

✅ How we built it

Built on Databricks using Delta Tables, LangChain, and Lakehouse Apps. Data was ingested from Minilabs datasets, embedded into a vector store, and connected to an LLM-powered agent that handles Q&A, summarization, and benefit recommendations.

✅ Challenges we ran into

Aligning unstructured and structured data for consistent summarization
Creating meaningful embeddings from complex healthcare benefit tables
Managing conversational accuracy with limited training data

✅ Accomplishments that we're proud of

Seamlessly integrated NLP and vector search over 69+ datasets
Delivered a working AI agent with contextual understanding in just a few hours
Enabled smart benefit summaries and service lookups via simple chat interface

✅ What we learned

Strong schema design and prompt tuning are critical for healthcare-related GenAI
Combining RAG with vector search significantly improves response quality
Real-world datasets bring unexpected variability, needing dynamic agents

✅ What's next for WellnessNavigator Agent