AICHE +Jira Integration

The short answer: open any Jira ticket, click into the description or comment field, press ⌃+⌥+R (Mac) or Ctrl+Alt+R (Windows/Linux), speak for 40-60 seconds, and AICHE inserts the formatted requirements or updates.

Jira tickets are where work gets defined. A vague ticket with "fix the login bug" sends a developer down a rabbit hole of guessing what's actually broken, which browser, which flow, what the user saw. A detailed ticket with reproduction steps, expected behavior, and technical context gets fixed in one pass. The problem is that writing a proper ticket with acceptance criteria, edge cases, and context takes 15-25 minutes. Most people don't have that time, so tickets ship half-baked, and the real cost shows up in wasted development cycles.

1. Open Jira in your browser or desktop client.
2. Create a new ticket or open an existing one from the backlog.
3. Click into the description field, or scroll to the comment box if you're updating a ticket.
4. Press your AICHE hotkey (⌃+⌥+R on Mac, Ctrl+Alt+R on Windows/Linux) to start recording.
5. Speak the full context. Include what's broken or what you want built, who it affects, how to reproduce it, and what "done" looks like.
6. Press the hotkey again. AICHE transcribes your speech, formats it, and inserts the text into the active field.
7. Add story points, assignees, sprint assignment, and component labels using Jira's sidebar fields.

Writing User Stories with Acceptance Criteria

The hardest part of a user story is not the "As a user, I want..." line. It's the acceptance criteria underneath. A PM who just finished a customer call has the full picture in their head: the user's frustration, the edge cases they mentioned, the integration that's breaking. That mental model decays fast. Within an hour, half the details are gone. Voice captures that model while it's still fresh. Press your hotkey and speak through the entire story, including the happy path, the error states, and the boundary conditions. Enable Content Organization in AICHE settings and your spoken stream gets structured into clear acceptance criteria sections that developers can work from directly. What used to be a 20-minute writing exercise becomes a 90-second dictation.

Bug Reports That Actually Get Fixed

The gap between a useful bug report and a useless one is reproduction steps. "Login is broken" tells the developer nothing. "Login fails with a 403 when using SSO through Okta on the staging environment after the session token expires" tells them exactly where to look. When you're staring at the bug, you know all of this. But by the time you open Jira, click New Issue, select the bug type, and start typing, you've already lost the thread. With AICHE, you stay on the broken page, press your hotkey, and speak the reproduction steps while you're looking at the error. The observed behavior, expected behavior, browser version, environment, and any console errors, all captured in real time. Paste a screenshot after the dictation and the developer has everything they need.

Sprint Retrospective Comments and Epic Descriptions

Retro comments typed during the meeting are usually short and vague because people are trying to listen and type at the same time. Voice changes that dynamic. When it's your turn to share, press your hotkey and speak naturally about what went well, what didn't, and what should change. AICHE's Message Ready feature polishes your spoken thoughts into clear, professional commentary. The same approach works for epic descriptions, which often sit empty because nobody wants to spend 30 minutes writing a summary of work that spans multiple sprints. Dictate the epic's goals, scope boundaries, success metrics, and key dependencies in two minutes. The epic becomes a useful reference instead of a blank placeholder.

Heads-up: Jira's rich text editor preserves your text after dictation, but it won't auto-format into bullet points or headings. After AICHE inserts the text, use Jira's toolbar to add structure. If you enable Content Organization, AICHE pre-structures the output so less manual formatting is needed.

Pro tip: before dictating a user story, say the ticket title first, then pause briefly and speak the description. This gives you a natural mental break between the summary and the details, and the resulting text reads more clearly.

Result: a backlog grooming session where you'd normally write 5 tickets in 2 hours produces 5 fully detailed tickets in 30 minutes, each with acceptance criteria, edge cases, and technical context that developers can start working from immediately.

Do this now: open a Jira ticket that has a vague or empty description. Press your hotkey and dictate the full problem statement, expected behavior, and acceptance criteria while the context is still in your head.