Programme Content
Modern AI Stack and Agentic Systems
Why AI matters for working professionals
Evolution of AI systems
Handling unstructured text input
How LLMs and agents became the new AI workflow stack
AI Systems Architecture and Data Foundations
Data fundamentals for AI workflows
ML concepts you must know for real use cases
What makes an AI system work end-to-end
How ML, LLMs, retrieval, and agents connect in real systems
AI Workflow Engineering and Application Design
Understanding AI workflow building blocks
Working with APIs, structured inputs, and outputs
Handling unstructured text and documents
Organising workflows using templates and reusable components
Applied Machine Learning for Production Use
Supervised and unsupervised learning basics
Building and validating small ML models using standard tools
Evaluating model behaviour with simple metrics
Exporting and reusing models in practical workflows
LLM Engineering and Prompt Design
Using LLMs safely and efficiently for workplace tasks
Prompt patterns for reasoning and structured outputs
Understanding context windows and token limits
Using LLMs for summaries, transformations, and workflow automation
Retrieval Systems and RAG Architecture
Document loading, chunking, and preprocessing
Embeddings and vector search basics
Connecting retrieval to prompts for reliable outputs
Building a simple RAG workflow end-to-end
Agent Orchestration and Tool-Integrated AI
What tool calling means and how it works in practice
Designing controlled tool inputs and outputs
Building multi-step agent workflows
Agent output checking and validation patterns
AI Application Layer and Deployment Fundamentals
Turning workflows into usable internal tools
Simple interfaces for teams and stakeholders
Environment setup and dependency basics
Introduction to monitoring, logging, and observability
AI Evaluation, Reliability, and Guardrails
Evaluating AI outputs for accuracy and reliability
Improving retrieval and prompting strategies
Error handling and fallback mechanisms
Basic safety, guardrails, and controlled execution principles
Capstone: End-to-End Agentic AI System
Participants build an end-to-end Agentic AI solution that includes:
LLM usage and workflow logic
Retrieval logic (RAG) where relevant
Tool-based workflow or agent orchestration
A usable interface or internal tool output
Clear documentation and demonstration
