talks about being a good AI Coach
summary
Claim: being a good AI coach means giving the model the right context and toolset, not vague prompts. The concrete workflow is prompt engineering then retrieval augmented generation then fine‑tuning, using system prompts, examples, embeddings or APIs (vector DB optional), and function calling to let agents run tools. Demos use LangChain/LangSmith and GitHub Models Marketplace and show a chatbot that writes a jingle and an agent that returns current time and weather, proving the approach works end to end. Test outputs against a gold standard.
AI-written summary · about
chapters (8)
- 0:00Being a good AI Coach
- 1:49Sports coaching to AI agents
- 3:38AI agents and generative workflow
- 8:39Fuzzy input fuzzy output
- 13:19Strategies for better AI results
- 19:04Building AI agents with function calling
- 24:31Adding security to AI agents
- 25:35AI versus traditional software
related talks
10mMastering the Art and Science of AI Communication
31mDesigning AI Agents with the Model Context Protocol: From Answers to Actions
27mEnhancing AI-Driven Quality Assurance
36mFrom Automation to Augmentation: AI Coaches That Amplify Expertise
24mAI as a Teammate: Redefining Agile for Human–AI Collaboration
24mThe truth about AI development
01h 39mMastering LLMs: From Prompt Engineering to Agentic AI
19mBeyond agents: AI as a creative partner