LLM Parameters: Model Weights vs Generation Settings
Teach the six building blocks — tokens, tokenization, context window, parameters, pre-training, inference — then use them to understand temperature, top_p, and max_tokens.
Coding concepts explained simply — the long-form companion to @prashant.code.
LLM Is Not an Agent
An LLM predicts the next token. An agent observes, decides, acts, and loops. Here's why confusing the two leads to broken demos and overbuilt systems.
How Large Language Models Work: A Visual Guide
Tokens, API calls, message roles, generation settings, tool calling, streaming, and limitations — a practical reference for building with LLMs.
Architecture of LangChain & LangGraph
Follow one request from the API call through state, tool calling, conditional routing, and evaluation — the path the 31 phases are actually describing.
31 Phases to Learn LangChain & LangGraph
The complete roadmap from my latest Instagram reel — every phase from LLM fundamentals to advanced agentic systems, explained.