🤖 AI Chatbot for Health Research using RAG and LLM

🧠 Overview

As part of my internship at PT. Inti Utama Solusindo, I worked as a Machine Learning Engineer and developed an AI-powered chatbot aimed at assisting health researchers in retrieving fast, accurate, and contextual information. The system utilizes Retrieval-Augmented Generation (RAG) and Large Language Models (LLMs) to generate intelligent responses based on structured knowledge extracted from open-access health journals.

🔍 Key Contributions

• 🧾 Data Scraping & Structuring
  Extracted and organized data from open-access health journals to form a clean, structured knowledge base.

• 🧠 AI & NLP Integration
  Applied state-of-the-art NLP and LLM techniques to improve chatbot understanding and response accuracy.

• 🔁 RAG Implementation with LlamaIndex
  Integrated the LlamaIndex framework to implement RAG for context-aware information retrieval.

• 🌐 Backend Integration with Django
  Developed a Django-based backend for serving chatbot APIs and managing response flow.

• 📊 Database & Performance Monitoring
  Utilized PostgreSQL for efficient data storage and optimized query performance. Used analytics to track interaction metrics and enhance system accuracy.

• 🤝 Cross-Functional Collaboration
  Collaborated with medical experts and data scientists to refine chatbot output and maintain response quality.

✅ Impact

• Improved access to reliable research information for health professionals and academics.
• Enabled domain-specific chatbot functionality tailored to the healthcare sector.
• Maintained up-to-date and relevant responses through periodic knowledge base updates.

🧰 Tools & Technologies

• Python, Django, PostgreSQL
• LlamaIndex, LLMs, NLP
• BeautifulSoup, requests
• Retrieval-Augmented Generation (RAG)

This project is a real-world application of AI and machine learning developed during my internship at PT. Inti Utama Solusindo, demonstrating how technology can support research and knowledge access in the healthcare industry.