Agentic AI Engineering with Python — Build Autonomous Agents Training Course

Core Concepts of Agentic AI

• Defining autonomous agents: classifications and core definitions
• The agent loop: the cycle of perceiving, deciding, acting, and observing
• Establishing design patterns for agent scope and responsibilities

Python Ecosystem and Agent SDKs

• Leveraging LangChain and comparable SDKs to initiate agent development
• Managing async programming, task queues, and subprocesses
• Implementing packaging, virtual environments, and reproducible development workflows

Connecting to External Tools and APIs

• Crafting tool interfaces and safe invocation patterns
• Establishing connections with web APIs, databases, and internal services
• Handling credentials, secrets, and enforcing least-privilege access

Managing Memory, State, and Context

• Utilizing short-term context windows and prompt engineering strategies
• Architecting long-term memory systems using Redis, vector stores, and retrieval augmentation
• Maintaining consistency, applying caching strategies, and ensuring memory hygiene

Orchestrating Multi-Step Workflows and Planning

• Implementing action chaining, subagents, and task decomposition
• Comparing planning algorithms with heuristic orchestration
• Managing failures, retries, and compensatory actions

Ensuring Safety, Testing, and Observability

• Developing threat models, conducting red-teaming, and sanitizing input/output
• Executing unit, integration, and end-to-end tests for agents
• Implementing logging, metrics, tracing, and alerting for agent behavior

Deployment, Scaling, and Agent MLOps

• Utilizing containerization, CI/CD pipelines, and rollout strategies
• Controlling costs, applying rate limits, and optimizing resources
• Overseeing monitoring, governance, and operational playbooks

Recap and Future Directions