Curriculum
Day 1 - Foundations of RAG AI Agents in Science & Healthcare
Goal: Understand RAG AI Agents concepts and build a basic Q&A system for scientific and clinical documents
Topics:
- Large Language Models (LLMs): Pre-training and limitations
- Retrieval-Augmented Generation (RAG): Enhancing accuracy with domain data
- Agentic AI
- ReAct framework
- Data Handling: Geoscience datasets (seismic, well logs, GIS layers) and healthcare datasets (EHRs, PubMed articles, imaging reports)
- Data Privacy & Compliance: HIPAA, de-identification, and secure scientific data handling
Hands-on Labs:
- Setting up a pre-trained LLM (GPT, LLaMA, T5)
- Uploading domain-specific datasets (scientific papers, clinical reports, geoscience logs)
- Implementing a basic Q&A retrieval pipeline
Day 2 - Building Domain-Specific AI Agents
Goal: Implementing Agentic patterns
Topics:
- Indexing specialized datasets (geoscience reports, medical literature)
- Vector Databases (FAISS, Pinecone, ChromaDB) for efficient retrieval
- Semantic Search vs. Keyword Search: Optimizing for technical terms
- Domain Adaption: Fine-tuning AI responses with scientific / clinical notes
Hands-on Labs:
- Building a Q&A assistant for medical literature and clinical notes
- Enhancing retrieval accuracy for industry-specific terminology
- Project Assignment
Day 3 - Deployment, Ethics, Applications and Networking
Goal: Deploy a functional AI assistant and explore real-world applications in geoscience and healthcare
Topics:
- Deployment Strategies: Local vs. cloud-based, API integration
- Cybersecurity and compliance in healthcare & scientific AI deployment
- Scaling retrieval systems for large datasets (e.g., multi-hospital records, global geological surveys)
- Ethical AI: Reducing hallucinations, ensuring accuracy, responsible use in healthcare decision-making
Hands-on Labs:
- Deploying a chat-based assistant for scientific & medical queries
- Testing retrieval and summarization on seismic surveys and clinical case studies
- Final Project: Demo of a domain-specific RAG AI agent
Deliverables
- Hands-on project in both geoscience and healthcare contexts
- Final deployed AI agent demo
- Certificate of Completion