From Business Intelligence to Artificial Intelligence

Week 1

The AI Landscape

• Set the foundation: what chatbots and agents actually are, what Claude Desktop can reach in each of its modes, and the prompting habits that separate reliable results from merely plausible ones.

AI Tools

• Work through the four tools Claude uses on your behalf — running code, searching and fetching the web, driving a real browser, and working at the command line — and the data mistakes that produce wrong numbers silently.

Week 2

Skills, Plugins, and Apps

• Turn a repeated task into a skill, share it with your team through a plugin marketplace on GitHub, and build web apps that produce analysis on demand.

AI Agents and Security

• Connect to your data through MCP connectors, build a custom agent with the Claude Agent SDK, and see where the risks come from: what a contract with a model provider does and does not cover, prompt injection, and data exfiltration.

Week 3

Deep Dive into Agents

• Build and run working agents in containers — a research agent, a data agent, and a multi-agent system whose architecture makes exfiltration impossible rather than merely unlikely — then red-team them on synthetic data.

Make Your AI Smarter

• Teach AI your own documents with retrieval, give it durable memory with a knowledge graph, build a secure personal assistant over your mail and calendar, and run smaller or fine-tuned models at a fraction of the cost.

Week 4

Decisions and Predictions

• Use AI for optimization, competing objectives, simulation, and prediction — and learn to verify the answers rather than trust them.

Capstone

• Pitch your AI implementation plan and get peer feedback.