Before the morning meeting, piece together a dozen tickets
Before each morning meeting, the team lead reviewed members' tickets one by one. Brief records meant searching emails and notes to understand progress before deciding how to help. Wade and Jane of Team 6 started with this routine, aiming to give team leads direct access to the key points.
Start with a small tool you will use yourself
The team considered several improvement directions during the workshop, but connecting too many systems created obstacles. They chose one everyday problem and built a local page the team lead would use personally.
The lead described inputs and expected outputs to AI in natural language: find open tickets, consolidate time in status, summarize the situation and make an initial assessment of whether the customer needs a reply. As needs emerged, the lead asked AI to add or modify features.
The page appeared, but some buttons did nothing
The first version had a UI, but ticket counts did not refresh immediately and some buttons had no function. The lead explained each expected interaction and result, then asked AI to fix it. The tool improved gradually through requirements and feedback rather than working perfectly on the first attempt.
The page now starts with the total ticket count. Individual tabs show tickets sorted by time in status, their situation and progress, and AI reply prompts. The lead can identify people and issues needing help before confirming them at the morning meeting.
Records help colleagues and give AI a reliable basis
Another key practice was asking members to consistently record the situation, progress and next step. The lead explained that this helps later review and leave handovers. Those same records, together with emails, also let AI summarize ticket highlights.
AI only provides initial extraction and prompts. If a summary differs from the lead's understanding, the lead checks with the person at the morning meeting and makes the judgment. The lead estimates preparation fell from about 30 minutes to 5–10 minutes. This is an individual usage estimate, not a formal measurement, and has not been validated with other leads. Saved time goes toward organizing the day's tasks and reviewing reminders.
Before and after: how the work changed
Compare the same work steps to see the roles of tools and people.
| Work step | Before | After |
|---|---|---|
| Pre-meeting review | Open PSA and review members' tickets individually, searching emails and notes when necessary. | Start with the total count, then use individual tabs to review tickets sorted by time in status, situation and progress. |
| Deciding whether to reply | The lead reads the ticket and decides whether a reply or assistance is needed. | AI extracts highlights and provides initial prompts for the lead to verify; differences are checked with the person at the morning meeting. |
| Preparation time | Preparing roughly 10–20 tickets for the morning meeting takes about 30 minutes. | The overview reduces preparation to about 5–10 minutes, freeing time to organize the day. These are the lead's personal estimates and have not been validated with other leads. |
| Work focus | Spend time gathering and piecing together information before considering how to guide members. | Identify people and tickets needing help first, then focus discussion on what needs clarification. Consistent situation, progress and next-step records also support review and handover. |
Applying this experience to your own work
The comparison uses the lead's personal time estimates, not results validated with other leads. Installer and environment testing are still underway, with deployment support from the AI team planned. Other teams can start with one meeting and one user, clarify the information and judgments needed, then verify each feature.