LLMs for Web Developers
summary
Walks through concrete patterns web devs can code today: basic prompt inferencing, RAG with a vector DB (example uses 200-token chunks, overlap, embeddings), and tool-calling/agents that plan and invoke APIs. Recommends LangChain and Chroma, shows ~30–80 lines of JavaScript to call models, uses TensorFlow GPU for local embeddings, demos with llama 3.2 and IBM/OpenAI endpoints, notes token costs (~$5 per million tokens) and that a hand-rolled agent example sometimes fails.
AI-written summary · about
chapters (10)
- 0:00Introduction to LLMs
- 6:26AI's impact on coding
- 9:32AI Engineer role
- 14:27Building AI solutions stack
- 15:30Prompt engineering basics
- 23:39Context and model cards
- 24:00Retrieval augmented generation
- 34:14Tool calling and agents
- 46:51Demo and resources
- 48:37Final thoughts and Q&A
related talks
36mLLMs and AI agents demystified
48mGoing Beyond Prompt Engineering
01h 59mLLM applications with LangChain
01h 34mA Practical Starter's Guide to building LLM based projects
27mAI Demystified Unveiling the Intelligence Behind the Machine
40mTechniques to avoid common LLM pitfalls
37mExpanding the Capabilities of LLMs Through Community Collaboration
01h 39mMastering LLMs: From Prompt Engineering to Agentic AI