All products

Find product improvement ideas with AI

Generate genuinely different directions for one product problem and inspect the evidence behind the shortlist.

You can find product improvement ideas with AI when a team has a recurring problem but keeps proposing variations on its first solution. Returning customers might struggle to locate previous orders. A search box is one possibility, but the order history, receipt or repeat-purchase journey could each offer a different route. The ce-ideate skill explores several directions before narrowing the list.

It can draw on authorised project material, earlier decisions and optional research tools. Existing research matters because a proposal that sounds promising may repeat an experiment the team has already tried. The output leaves room to distinguish evidence from a hypothesis worth testing.

What it does

  • Explore alternatives to a feature request already on the backlog.
  • Compare ideas grounded in different parts of a customer journey.
  • Surface relevant decisions from authorised project records.
  • Identify where the supporting evidence is thin.
  • Select one direction for a more detailed requirements conversation.

Install and use

  1. Use Compound Engineering in a supported agent and define which project sources it may access. Connected Slack research can search accessible private channels as well as public ones.
  2. Describe the problem with a concrete example and name any information that must stay internal.
  3. Review the proposed directions, the available evidence and the most useful question to answer next.

FAQ

Source

Made by Every. Original skill: ce-ideate, published on GitHub under the MIT licence.

View on GitHub