Skip to main content
Use this page when AI should help with real Custory work and the result needs to stay attached to the journey, item, or conversation that produced it.

​
What this page covers

This page explains how AI participates in Custory as part of the workspace instead of as a detached prompt box.

​
Why it matters

For small teams, the biggest AI problem is usually not raw output quality. It is the work getting split across too many places, so nobody can easily tell what happened or why.

​
What working with AI looks like in practice

Treating AI as a workspace member gives Custory four useful qualities:
  • a consistent identity inside the workspace
  • mention-based collaboration
  • async work with visible outcomes
  • a visible record of what happened

​
A consistent identity

Because AI is represented as an actual member, it fits the same teamwork model as the rest of the workspace. That means AI requests can live:
  • near the journey
  • near the item
  • near the team discussion
instead of being copied into an external prompt box that no one else can trace later.

​
Mention-based collaboration

Custory AI can be mentioned directly where the work is happening. That is useful when the team wants help with:
  • reviewing a journey for gaps
  • summarizing evidence
  • drafting a follow-up task
  • reframing an opportunity
  • suggesting a cleaner summary for Slack or Discord

​
Async work with visible outcomes

Not every AI task should behave like instant autocomplete. Some jobs require real processing time:
  • reviewing larger context
  • drafting follow-up output
  • running a multi-step task
  • working with connected tools
Custory supports that by letting AI work asynchronously and return a visible result afterward.

​
AI and building blocks

Custory AI can work directly with reusable content in building blocks. That includes:
  • creating repository items
  • attaching existing blocks to journeys
  • removing a block from a journey without deleting the reusable source item
  • updating shared insights, opportunities, solutions, metrics, or touchpoints from repository context
This is useful when the team wants AI to update the shared record once and place it on the right journeys afterward.

​
AI and automations

AI also helps inside automations when the output needs to reflect the journey, item, or source context instead of following a static template. Examples:
  • weekly journey summaries
  • repository refreshes for shared metrics or insights
  • issue descriptions written from real customer context
  • journey or building-blocks updates after a GitHub or analytics change

​
What AI is not replacing

Custory AI can help the team move faster, but it does not replace:
  • product judgment
  • customer validation
  • priority calls
  • clear ownership
It works best when it reduces mechanical work and makes the next decision easier to review.

​
AI mistakes to avoid

The quality of outputs improves when the journey, items, and linked evidence are already in decent shape. Give AI something real to work from instead of expecting it to invent clarity.
AI can draft, summarize, and suggest, but the team should still review anything that changes real product work. The best pattern is to use AI for speed and humans for judgment.
The more detached the request is from the journey context, the weaker the result usually becomes. Keep the prompt as close as possible to the item, journey, or decision it is meant to support.

​
What strong AI usage looks like

Good AI usage in Custory feels like:
  • less manual synthesis
  • better follow-up drafts
  • faster review preparation
  • clearer outputs that stay tied to the source context

​
Next step

Read Workspace memory if you want AI to retrieve relevant prior context. Read Slack/Discord threads if your team collaborates with AI in chat.