Back to Portfolio Try Live App

Medical Summary Agent

AI-powered research agent for accessible, evidence-based healthcare information

Python Streamlit OpenAI API PubMed API AI Agents

The Problem

AI Hallucinations in Healthcare

ChatGPT and similar AI models often "hallucinate" medical information—generating plausible-sounding but factually incorrect health advice. This is dangerous when people rely on AI for medical decisions.

Inaccessible Medical Research

The latest peer-reviewed medical research is often locked behind paywalls, written in technical jargon, or scattered across databases—making modern medicine inaccessible to everyday people.

The Solution

The Medical Summary Agent is an AI-powered tool that bridges the gap between cutting-edge medical research and accessible healthcare information. It combines the power of AI with the reliability of peer-reviewed science.

1. Query PubMed Database

The agent searches PubMed—the world's largest database of peer-reviewed medical literature—for the most recent and relevant research papers on your health query.

2. AI-Powered Analysis

Using OpenAI's API, the agent reads and synthesizes multiple research papers, extracting key findings and translating complex medical terminology into clear, understandable language.

3. Evidence-Based Summaries

The agent generates summaries grounded in real scientific evidence, complete with citations to the original research papers—eliminating hallucinations and ensuring accuracy.

Technical Implementation

AI Agent Architecture

  • • Built with Python and Streamlit for rapid deployment
  • • Implements agentic workflow with tool-calling capabilities
  • • Uses retrieval-augmented generation (RAG) pattern
  • • Integrates OpenAI API for natural language processing

PubMed Integration

  • • Queries NCBI's PubMed API for peer-reviewed articles
  • • Filters for recent, high-quality research papers
  • • Extracts abstracts and metadata for analysis
  • • Provides direct citations to source material

Impact & Goals

Accessible Modern Medicine for All

This project embodies the principles of effective altruism by democratizing access to medical knowledge. Healthcare information should not be a privilege—it should be a right.

🎯

Accuracy

Grounded in peer-reviewed research, not AI hallucinations

🌍

Accessibility

Free, open access to the latest medical science

đź’ˇ

Clarity

Complex research translated into plain language

AI Safety & Ethics

This project demonstrates responsible AI development by addressing key safety concerns:

Grounding in Truth: By anchoring AI responses in peer-reviewed literature, we prevent hallucinations and misinformation.
Transparency: Every summary includes citations, allowing users to verify information and read the original research.
Human-Centered Design: The tool empowers users to make informed decisions, not replace medical professionals.

Try the Medical Summary Agent

Experience how AI can make medical research accessible while maintaining accuracy and reliability.