Agentic AI Engineering Course

Source: https://academy.towardsai.net/courses/agent-engineering

Your Path to Agent AI for Production

Become an Agent AI Engineer, create systems ready for real-world use, and build a portfolio of AI agents that showcase your skills.

What You Will Learn:

• When an Agent is Needed and When Not – We will teach you how to avoid the three main pitfalls developers face when creating agents.
• Agents Work Locally but Break in Production – You will build two fully-fledged production agents with deployment, observability, and pipelines.
• Solving Complex Tasks – Create an autonomous research system and a workflow for content creation with automated quality assessment and tool integration.
• Mastery of Decision Making – Learn to choose between deterministic routing and agent autonomy, determine when to keep a human in the loop, and design scalable prototypes for production.
• Skills that Don’t Become Obsolete – We teach fundamental system design, not just contemporary frameworks. Graduate with two fully deployed agents and engineering experience that will allow you to build any AI systems.

What You Will Create:

• Research Agent – Autonomous data collection and structuring from the web, GitHub, and YouTube, iterative research cycles, and tool integration for analysis.
• Content Creation Workflow – Generation of texts, diagrams, and images, automated quality checks, context-style management, and a production pipeline.

Agent Architecture:

• Reliable and scalable workflows
• Custom evaluation and monitoring systems
• Deployment in Docker and cloud environments
• Integration with Cursor, Claude, and other tools

Course Outcome:

• Two production agents in your portfolio
• Understanding of fundamental principles that will outlast trends

Who This Course is For:

• Advanced Python users (functions, classes, API)
• Familiar with LLM and basic prompting (OpenAI/Claude)
• Understanding Docker and basic deployment principles
• Ready to learn through practice rather than watching video lessons

Password: cms.ddpanda.org