Fundamentals of AI agents

Here you will lay the theoretical foundation and set up your technical environment so that you are ready to work in the subsequent webinars. 

• Deep dive understanding: Why agents need external tools (tool use) and long-term memory (memory, e.g., Pinecone) to solve complex tasks

• Glossary & Tech Stack: Detailed explanation of key terms and tools (LangChain, Make, n8n, Context, vector databases). 

• Low-code vs. code: A decision-making aid for getting started. Analysis of when low-code (n8n/Make) is sufficient and when engineering (LangChain> Python/TS) is necessary. 

• Setup: Instructions for setting up the necessary accounts so that everything is ready to go for the next webinar. 

• Knowledge check & homework for webinar 2:Think of an agent use case for the second webinar. 

The focus is on the concept ("the agent's brain") and the economic perspective.

Strategy & Use Cases: 

• Decision matrix: Differentiation in practice — is my case an assistant, a workflow, or an agent? 
• Business Model Canvas for agents: Development of goals, required tools, contextual data, and ROI calculation. Validation of your own business case. 

Context Engineering (Hands-on): 

• Context > Prompt: Why engineering context is more important than "prompt magic." 
• Live building: Together with the trainer, you as a learning group will build the "brain" of an agent (system prompts + context data) live. 
• Structuring: Write system prompts and define what knowledge the agent needs. 

Deepening of concepts and preparation for the big agent construction day in Webinar 3. 

• The "Good Context Checklist": Criteria catalog for robust context. You check your designed use case for data completeness. 
• Enterprise readiness: A critical examination of the stability of solutions. Distinguishing between hype demos and stable enterprise solutions. 
• Fine-tuning the process: You prepare your company's internal process so that it can be technically implemented in the next webinar (create a flow chart, review API documentation). 

The big implementation day. We build real agents and make them ready for operation. 

Deep Dive Building: 

• Hands-on group work: Building an agent depending on skill level (e.g., research agent, RAG document agent, or Workflow Plus agent). 
• Integration: Connecting LLM (brain) and tools (hands). The trainer provides direct support in the breakout rooms if problems arise. 

Production Readiness & Governance: 

• Reality check: Dealing with loops, hallucinations, and unclear tools. 
• Safety first: Implementation of kill switches, human-in-the-loop governance, and logging. 
• Responsibility: Data protection, employee vs. agent responsibility, and ethical issues. 
• Roadmap: Outlook on upcoming technologies (e.g., Manus, Antigravity) and strategies for integration into the team ("Build with Agents vs. Act with Agents").