Skip to content

AI estimation

AI that helps your team estimate, not replace them

Story-aware suggestions, automatic decomposition, completeness checks, and an AI facilitator that surfaces disagreement. Always optional — humans cast the votes that count.

Start free — 7 days TeamSee credit pricing
PROJ-218User authentication flow
JKMPASLN2 of 4 voted
AI
AI Analysis

Analyzing: User authentication flow

8

Confidence

Medium confidence

Similar stories

PROJ-142OAuth login
8 pts
PROJ-89Session management
5 pts

Story Decomposition

  • OAuth provider setup
    3 pts
  • Token refresh logic
    3 pts
  • Session expiry UI
    2 pts
Total:8 pts
5
Credits on free
50
Credits on Pro
200
Credits on Team
30
Trial credits

Why it works

Why AI here, not somewhere else

Estimation is the perfect AI use case: structured input, bounded output, and a human team that decides whether the suggestion lands. AI receives story text and your deck. It returns a card from your own deck with a one-line reason. You vote.

Decomposition

Split a 13 into three shippable stories

If a story comes back as a 13 or higher, you almost certainly need to split it. AI Decomposition proposes the split before the team wastes a round re-voting.
  • AI takes the oversized story and proposes 2–4 sub-stories with individual estimates.
  • Edit the sub-stories inline, accept what works, discard the rest.
  • Write accepted sub-stories back to Jira or Linear with one click.
  • Each decomposition uses 2 credits — the cheapest way to fix a bad estimate.

Risk radar

Catch the risks a tired scrum master would miss

The stories most likely to blow an estimate are the ones that look straightforward. Risk Radar surfaces the hidden complexity before the team commits.
  • Highlights delivery, integration, and estimation risks before voting starts.
  • Flags third-party API dependencies, ambiguous wording, and missing edge cases.
  • Severity levels (amber / rose) so the team knows what to discuss first.
  • Each risk-radar run uses 2 credits and runs at story import time.

Readiness

Score story completeness before the vote

Before voting starts, AI scores how ready a story is to estimate. Refinement happens before the team burns time on a bad estimate.
  • Checks for acceptance criteria, assigned owner, and obvious unknowns in seconds.
  • Stories scoring under 70% are flagged before voting starts.
  • Prevents wasted rounds on stories that aren't ready to estimate.
  • Uses 1 credit per story — or 1 credit for the whole batch.

Facilitation

AI Discussion Facilitator (Team plan)

When votes split, the Facilitator suggests targeted follow-up questions and summarizes the disagreement. Available as a streaming chat-style flow — keeps the human moderator in charge.

Credits

How credits work

Each AI action consumes a small number of credits. Plan credits reset monthly. Top-up packs (50–2,500 credits) never expire.

Free5 AI credits per month
Pro50 AI credits per month
Team200 AI credits per month
Enterprise1000 AI credits per month

Privacy

What the AI sees — and what it doesn't

  • AI receives: story title, description, the active deck, and your team's past story+estimate pairs.
  • AI never receives: source code, attachments, customer data, or personally identifiable information.
  • Every AI request is logged in your team settings — full transparency.
  • AI features can be disabled at user, room, or workspace level.

Try AI estimation in your trial

Sign up gets you 7 days of Team with 30 AI credits — enough to try every feature on a real backlog.

Start free — 7 days TeamSee credit pricing