Back to all work
AIHealthcareSaaS

BeBalanced

Intelligent medical assistant chatbot with persistent memory.

BeBalanced

Details are restricted under non-disclosure agreements.

Client NDA Protected
Year
2025
Client
CCRIPT Agency
Role
Senior Full Stack AI Developer
Status
Proprietary

Overview

The brief

Built an intelligent medical assistant chatbot using LangChain and a RAG pipeline backed by Postgres. Enabled users to describe health symptoms, receive guidance, and track medications. Patient test results and health histories are stored and retrieved contextually, supporting accurate, personalized responses across sessions.

The Challenge

Medical platforms often lack session-to-session continuity. Patients are forced to describe their medical history, chronic symptoms, and prescription schedules from scratch on every interaction.

Features

What we shipped

  • 01PostgreSQL RAG pipeline for medical database query matching
  • 02Contextual session-to-session memory persistence
  • 03Patient health history and test result semantic retrieval
  • 04Medication tracking and personalized guidance

Tech Stack

Next.jsNode.jsPostgreSQLLangChainRAGPython

Impact & Performance

<150ms
Search Latency
98.4%
Response Accuracy
RAG & LangChain
Pipeline

Next

See all work →