It was a pleasure to host
@augvcor for a hands-on workshop focused on building and evolving real-world AI agents. The session offered a practical, end-to-end walkthrough—from core concepts to a fully functional, tool-enabled AI assistant.
Key takeaways from the workshop:
➡ Understanding how AI agents and large language models work in real-world scenarios
➡ How agents maintain context, reason over inputs, and decide between direct responses and tool usage
➡ Introduction to GitHub Codespaces as a fully cloud-based development environment
➡ Exploring multiple ways to create and launch a Codespace
➡ Forking the provided GitHub repository to experiment freely with agent code
➡ Navigating the project structure using the repository’s step-by-step guide
➡ Creating and configuring GitHub access tokens for secure resource access
➡ Setting up a dedicated Python virtual environment for dependency isolation
➡ Installing required frameworks, including Chainlit for chat UI and LangChain for agent logic
➡ Building a working chatbot using Chainlit with a web-based chat interface
➡ Enhancing the chatbot by integrating an external weather API
➡ Understanding the agent’s decision-making flow between API calls and language model responses
➡ Customizing chatbot behavior to create a more natural, conversational assistant
➡ Gaining hands-on experience in evolving a basic LLM chatbot into a capable, tool-using AI agent
Thank you, Augustine Correa, for an engaging and practical workshop that enabled participants to build, experiment, and truly understand AI agents in action.
📍 Live moments from our tech event
#TechMang26 at
@itswrkwrk ✨
#Workshop #AIAgents #LLM #Chainlit #LangChain #GitHubCodespaces #HandsOnLearning #AIDevelopment #TechCommunity #KnowledgeSharing