Clinical Trials AI Recruiter (MedMatch AI)

Volunteer recruitment AI for clinical trials. Uses RAG (Retrieval-Augmented Generation) on trial-criteria PDFs + volunteer CSV to return fast eligibility counts, shortlists, and 2–3 sentence answers.

🔗 Live

• Website (Landing): https://ishaan0709.github.io/Techvital_clinical_trials/
• AI App (Streamlit): https://ishaan000999-pb9qvwjtdda7wdeqctz4fd.streamlit.app/
• Dev locally: http://localhost:8501

⚙️ Stack

Streamlit • Python • LangChain • RAG • FAISS • OpenAI • PyMuPDF (fitz) • pandas

🧠 How it works

1. Parse criteria PDF → extract: Age between 30 and 75 years, Non-Small Cell Lung Cancer (NSCLC), EGFR-positive (EGFR+), Stage III or IV.
2. Chunk text → FAISS embeddings; query with LangChain + OpenAI.
3. Filter CSV by age/gender/region/stage/biomarker; answer with concise summaries.

💬 Try in the Medical tab

• “How many eligible volunteers?”
• “Eligible EGFR+ Stage IV volunteers?”
• “Eligible females from Mumbai (show up to 5 IDs).”