What happens once training stops: frozen weights, prefill vs decode, sampling, the KV cache, and the serving tricks that make inference faster without changing what the model knows.
How SFT and preference tuning (RLHF/DPO) turn a pre-trained next-token model into a helpful assistant: same training loop, different data, thin layer that shapes behavior not knowledge.
What happens inside the weight tensors as an LLM trains: random initialisation, the four-step training loop (forward, loss, backprop, optimizer), and what each component learns as a next-token model.
Hyperparameters, vocabulary, and why a 7B model is 7B before day one, how the dials connect, where parameters live, and why the context window is a budgeted choice.
MCP is a valuable standard for AI tool interoperability. For regulated business agents, pin agent governance first (route contract, tools, permissions, context), then choose MCP only where it beats governed business APIs.
A 20,000-ft view of the LLM lifecycle and why understanding the four stages matters for enterprise architecture.