Microsoft Foundry & Python: Enterprise AI Agents on Azure

Descripción

Welcome to enterprise AI in 2026: the era of autonomous agents is here. Organizations today demand more than simple chatbots - they need intelligent systems that reason, retrieve knowledge, enforce policies, orchestrate tools, and execute real business workflows at scale. This hands-on, production-grade course teaches you how to build enterprise AI agents on Azure using Microsoft Foundry (the unified AI app & agent factory) and the Microsoft Agent Framework (the official 2026 successor to Semantic Kernel and AutoGen). This is the exact stack top corporations use for governed, scalable, and production-ready agentic AI. The Microsoft Foundry Advantage Microsoft Foundry is the "Agent Factory" for the modern enterprise. It unifies high-performance models (OpenAI, Anthropic, Llama, DeepSeek and more), centralized knowledge via Foundry IQ, and built-in governance in one secure environment. Paired with the open-source Microsoft Agent Framework, you will go from local Python prototypes to compliant production deployments on Azure. Course Curriculum at a Glance Phase 1: Foundations & Setup • Agentic AI vs. Traditional Automation: When to use which. • Microsoft Foundry Architecture: Understanding Hubs, Projects, and the Agent Service. • The Microsoft Agent Framework: Deep dive into the unified SDK for Python. • Environment Setup: Azure CLI, VS Code AI Toolkit, and Python SDK configuration. Phase 2: Building the Customer Support Agent • Foundry Playground: Rapid prototyping and agent configuration. • Tool Calling: Connecting your agents to real-time APIs and business systems. • Foundry IQ & Knowledge Retrieval: Implementing enterprise-grade RAG. • Responsible AI: Setting up Guardrails, Prompt Shields, and Grounding policies. • Front-End Integration: Building a Streamlit UI to interact with your agents. Phase 3: Multi-Agent Workflows & Orchestration • Agents vs. Workflows: The decision framework for enterprise architects. • Foundry Workflow Designer: Building deterministic, multi-step AI pipelines. • Agent Communication Patterns: Orchestrating "teams" of agents to solve complex tasks. • Observability: Using Traces and Application Insights to debug and monitor agent reasoning. Phase 4: The Capstone Project • Design & Architecture: Modeling a Support Triage system for a large organization. • Intelligent Routing: Using LLMs to categorize and route requests with high precision. • Autonomous Execution: Validating end-to-end production-grade agent workflows including human-in-the loop. What You Will Build • Full Customer Support AI Agent: An autonomous system with reasoning, multi-step resolution, and tool-calling capabilities. • Advanced RAG with Foundry IQ: Grounding your agents in enterprise data using Azure AI Search and Foundry’s intelligent retrieval layer. • Memory-Enabled Corporate Agents: Systems with persistent state that maintain context across complex, long-running business cycles. • Strict Enterprise Governance: Implementing AI guardrails, prompt shields, and Responsible AI controls to ensure safety and compliance. • Streamlit Front-End: A professional dashboard for agent interaction, testing, and real-time observability. • Capstone Project: An Automated Support Triage Workflow featuring intelligent ticket routing, prioritization, and autonomous execution. Who This Course Is For • Enterprise Developers & Architects leading the transition to agentic AI on Azure. • Cloud & Platform Engineers advancing from simple LLM calls to complex AI orchestration. • Python Developers who want to master the official, 2026-standard Microsoft AI stack. • Consultants & Technical Managers delivering autonomous business solutions for regulated industries. Why This Course Stands Out • 100% Current: Fully updated for the 2026 Microsoft Foundry rebrand and Agent Framework standards. • Enterprise-First: We prioritize security, compliance, and grounding—not just "happy path" demos. • Production Patterns: Learn the architectures required for deployment in large, regulated organizations. • Practical Python Integration: Focus on the Python SDK patterns used by professional AI engineers. Enroll Today The future of enterprise software is agent-driven. Join now to master Microsoft Foundry and the Microsoft Agent Framework, and position yourself at the forefront of the 2026 AI transformation.

Lo que aprenderás

• Build and deploy production-grade AI agents on Azure using Microsoft Foundry and Agent Framework
• Create enterprise AI agents with tool calling, advanced RAG, persistent memory, and strict grounding
• Design and orchestrate multi-step AI workflows – call them directly from Python
• Implement responsible AI: configure guardrails, prompt shields, and observability for compliant systems
• Develop a full Customer Support Agent + Streamlit UI for real-world interaction
• Build an automated Support Triage Workflow capstone – intelligent routing and autonomous execution

Requisitos

• No prior experience with Microsoft Foundry, Agent Framework, or advanced AI agents needed - we start from core concepts.
• Basic Python knowledge (helpful: functions, imports, APIs)
• Basic familiarity with LLMs (how they work, models, tokens) – no advanced experience needed
• Familiarity with cloud concepts (helpful but not required).
• An active Azure subscription (free tier sufficient for most exercises).