Building Autonomous AI & Data Agents with Python

From Static Prompting to Autonomous Multi-Agent Systems

The paradigm of artificial intelligence has shifted. The future belongs not to passive chatbots, but to autonomous agents: software systems that can formulate hypotheses, use external software tools, query production databases, self-correct code execution errors, and collaborate in multi-agent teams.

Curated by Dr Stylianos Kampakis, CStat (Chartered Statistician, The Alan Turing Institute partner), this flagship course teaches you how to build production-grade AI agents from first principles using Python, standard tool-calling protocols (including MCP), and multi-agent supervisory frameworks.

What You Will Master:

• ReAct Loop Foundations: Build a clean, transparent Reason+Act execution cycle without bloated third-party frameworks.
• Model Context Protocol (MCP) & Tools: Standardize agent interfaces for secure SQL database queries and local Python sandboxes.
• Multi-Agent Orchestration: Implement Supervisor-Worker, Router, and Generator-Critic patterns for complex workflows.
• Reliability & Guardrails: Prevent infinite execution loops, token budget overruns, and state explosion with deterministic circuit breakers.
• Production Deployment: Package and deploy an autonomous data research agent with full observability and telemetry.