Prompt engineering: definition

Prompt engineering is the craft of writing the instructions, examples, and context you give a model so it produces reliable, correctly formatted output for a specific task.

The same model can be useless or excellent on the same task depending on how you ask. Prompt engineering is being deliberate about that: stating the role, giving a couple of worked examples, defining the output shape, and spelling out the edge cases.

In production it matters less as a clever trick and more as discipline. A prompt that is versioned, tested against real inputs, and tightened over time is what keeps an automated step trustworthy.

Related terms

Large language model (LLM): A large language model is a neural network trained on huge amounts of text to predict the next token, which lets it write, summarise, classify, and answer questions across almost any topic.
Fine-tuning: Fine-tuning is further training a base model on your own examples so it learns a specific style, format, or task more reliably than prompting alone can achieve.
Hallucination: A hallucination is when a model states something false or invented as if it were fact, because it predicts plausible-sounding text rather than looking anything up.