AI Agents Course

Issued by: Hugging Face
Instructors Ben Burtenshaw, Sergio Paniego
Completed on: May 17, 2025

Course Overview: This course provides a comprehensive, hands-on journey into the world of AI Agents, covering everything from foundational concepts to advanced implementation and evaluation. The curriculum is designed to build practical expertise in designing, building, and deploying autonomous agents using cutting-edge tools and libraries. Through a series of hands-on labs, real-world assignments, and competitive challenges, I gained a deep understanding of agent architecture and its application to complex problems.

Key Skills Acquired:

• Agent Fundamentals: Understanding of core agent components like Tools, Thoughts, Actions, and Observations (TTAO).
• LLM Integration: Proficiency in using Large Language Models (LLMs) as the reasoning engine for agents, including message formatting and chat templates.
• Agent Frameworks: Hands-on experience with popular agent development libraries such as smolagents, LlamaIndex, and LangGraph.
• Advanced Techniques: Fine-tuning LLMs for function-calling, implementing agent observability, and performing rigorous evaluation.
• Practical Applications: Built and deployed agents for real-world use cases and participated in a competitive challenge to benchmark agent performance.
• Development Tools: Python, Hugging Face Hub (Spaces, Model Sharing).

Curriculum: The course is structured into several core units and specialized bonus modules:

1. Agent Fundamentals: Explored the theoretical underpinnings of AI agents and built a simple agent from scratch.
2. Frameworks: Dived into established libraries to understand how foundational concepts are implemented for robust agent development.
3. Use Cases: Applied learned concepts to build agents for practical, real-world scenarios.
4. Final Assignment: Developed a sophisticated agent for a selected benchmark, competing on a leaderboard to validate its performance.
   • Bonus Unit 1: Fine-tuning an LLM for Function-calling: Learned to adapt LLMs specifically for tool use.
   • Bonus Unit 2: Agent Observability and Evaluation: Gained skills in monitoring and assessing agent behavior and performance.
   • Bonus Unit 3: Agents in Games: Explored a fun and challenging application by building an agent to play Pokémon battles.

This course has equipped me with the theoretical knowledge and practical skills to build and deploy intelligent, autonomous AI agents capable of solving complex tasks.

You can view the certificate here.