AI Features in PM

Hubrix PM has three AI helpers built in: generating stories from a description, analysing an individual story, and planning a sprint. Each one uses the same AI providers as the rest of Hubrix and is grounded in your project's real data.

The AI story generator turning a short requirement into a set of draft stories
The AI story generator turning a short requirement into a set of draft stories

Generate stories from a prompt

On a project, open the AI generator and describe what you need in plain language — for example, "Build a customer onboarding flow with email verification and a welcome checklist."

Hubrix drafts a set of stories for you, each with:

  • a concise title,
  • a short description,
  • a suggested priority,
  • a story points estimate, and
  • up to two suggested labels.

You can ask for a specific type (task, story, or bug) and a count — up to 10 stories per request (the default is 5). Review the drafts, adjust anything you like, and keep the ones you want.

Analyse a story

Open a story and choose Analyze. Hubrix looks at the story together with the project context and your team's recent velocity, then returns suggestions:

  • a recommended priority (with the reasoning behind it),
  • a suggested assignee (with why they're a good fit),
  • a suggested story points estimate,
  • a note on risk, and
  • suggested tags.

The suggestions are advisory — nothing changes on the story until you apply it.

Plan a sprint

Open a sprint and choose AI Plan. Hubrix reviews your backlog against the sprint goal and your team's average velocity, then proposes a sprint plan:

  • the stories it recommends pulling in (by ref),
  • the total points of the recommendation (kept within your velocity),
  • a fit score from 0–100 indicating how confident it is,
  • a one-sentence sprint goal suggestion, and
  • any warnings about over-commitment or risk.

If you like the plan, click Apply Plan and Hubrix moves the recommended stories into the sprint (optionally applying the suggested point estimates). You stay in control: edit the recommendation before applying, or apply only part of it.

The sprint planner is only as good as your data. Keep the sprint goal filled in, estimate stories in points, and complete sprints so velocity is accurate — the planner uses all three.

What it costs

The PM AI helpers make AI calls, so they draw on your workspace credits like other AI features. See Billing → Credits for how credit usage works.

Was this helpful?