Large language model (LLM)
A large language model is a neural model trained on text to predict and generate language, often used as a reasoning component inside apps and agents.
How it works
Products wrap LLMs with prompts, tools, retrieval and guardrails. The model proposes; your system constrains. Evaluation measures task success, not vibes.
Example
An internal tool uses an LLM to turn messy notes into a structured ticket, validated against a schema before it hits the tracker.