Requirements elicitation uncovers what a person actually needs before implementation begins. Questions can explore purpose, audience, constraints, priorities, risks, inputs, expected outputs, and how a successful result will be judged.
An AI assistant can make the process conversational by asking one focused question at a time and suggesting plausible answer choices. The user still owns the decisions, especially where different answers would materially change scope or cost.
ELI5
Requirements elicitation turns a vague request into a clear description of what needs to be built. It asks about the goal, users, priorities, limits, risks, inputs, expected results, and how success will be judged before implementation begins.
For example, instead of immediately building a 'news app', an assistant might ask who it serves, which sources are allowed, how often it updates, and whether users need accounts. The answers prevent major assumptions from silently changing scope or cost.
Why not start building from the first idea?
Early clarification can reveal conflicting assumptions, missing users, weak business logic, and constraints that would otherwise force expensive rework.
How can suggested answers help requirements elicitation?
They reduce blank-page effort and show the kinds of decisions needed, while still allowing the user to reject or revise every option.
