Retrieval-Augmented Generation (RAG) Training in Australia

This course teaches Retrieval-Augmented Generation (RAG) end-to-end — an approach where a Large Language Model (LLM) dynamically retrieves relevant external data to generate contextually accurate responses rather than relying solely on its training set. Doing so enhances accuracy, relevance and traceability of AI outputs.

By the end of the program, participants will be able to:

• Understand RAG architecture and workflows
• Construct RAG pipelines from data sources to vector search
• Deploy scalable RAG services
• Evaluate and optimize performance metrics

This training is ideal for:

• Data Scientists & ML Engineers

• NLP Researchers

• Backend / Platform Engineers

• Product & Project Managers

• AI Technology Enthusiasts

Graduates of this training will be able to:

• Build and deploy RAG-powered AI applications
• Enhance LLM outputs with external knowledge
• Reduce hallucination risk
• Optimize RAG performance and accuracy
• Design robust, scalable solutions for modern AI systems

Introduction to RAG & Key Concepts

What is RAG — definition and motivations

Limitations of static LLMs

Retrieval + Generation pipeline

Information retrieval evolution and dense retrieval

RAG history and trends 

Data Sources, Chunking & ETL

Structured and unstructured data sources

ETL with modern tools

Chunking strategies for RAG systems

Metadata enrichment 

Embedding Design & Optimization

Choosing the right embedding models

Multilingual and domain-adapted embeddings

Quality optimization and fine-tuning

Embedding versioning 

Pipeline & RAG Architecture

Folder hierarchy for RAG systems

Continuous Integration & Delivery (CI/CD)

Performance metrics and evaluation

Validation automation 

Monitoring, Security & Compliance

Real-time observability

Data and model security best practices

Source citation and traceability

GDPR and AI compliance