Also known as: prompting · chain of thought · few-shot prompting
Prompt engineering is the practice of structuring input to a language model — through instructions, examples, formatting, and reasoning scaffolds — to reliably obtain the desired output.
Techniques that consistently help include supplying concrete examples of the desired output, requesting explicit step-by-step reasoning before an answer, specifying output format precisely, and stating what to do in ambiguous cases rather than leaving them undefined.
As models improve, elaborate prompting tricks have diminishing returns while clarity remains valuable. In production the durable work has shifted from clever phrasing toward context construction: deciding what information the model should see, which is a retrieval problem more than a wording problem.
Large Language Model
A large language model (LLM) is a neural network trained on vast quantities of text to predict likely continuations of a sequence, producing systems capable of generating coherent language, answering questions, and writing code.
Context Window
The context window is the maximum number of tokens a language model can consider at once, encompassing both the input prompt and the generated output.
Retrieval-Augmented Generation
Retrieval-augmented generation (RAG) is a technique in which relevant documents are retrieved from an external knowledge source and supplied to a language model as context, so that its response is grounded in specific retrieved material rather than parametric memory alone.
AI Evaluation
AI evaluation is the practice of systematically measuring model performance against defined tasks and criteria, using benchmarks, human review, or automated graders to quantify capability and detect regressions.