CURRENT CASE MAP
From implementation to results: see how work changes
Teams 1, 3, 4, 5 and 6 have shared complete stories. Start with the outcomes, explore how teams solved implementation challenges, and find approaches for your own work.
Compare the work before and after, then read the implementation story. Tool availability is stated separately; a published story does not mean the tool is available to everyone.
01Workflow connected
Order emails become drafts, then enter BC after human review
Team 1 | Vida (General Manager's Office) × Matt (Sales)
AI transcribes data and suggests historical item numbers for SA review. Order rules are documented, pending items are marked separately, and direct email intake reduces manual forwarding.
- Who can learn from this
- Sales assistants, sales support and colleagues who enter email or attachment data into systems.
- Before
- Read emails field by field, look up item numbers and enter data manually.
- After
- AI prepares drafts and suggests item numbers; people verify them before writing to BC.
Try this approach: start with a draft that can be verified; return multiple item-number candidates to a person for judgment.
Story available; no tool-access link is provided on this page yet.
03Guide in use
New hires follow setup guidance, reducing step-by-step assistance
Team 3 | CC (Administration) × Ian (Product)
Assistance per new hire fell from about 10–15 minutes to 3–5 minutes (team estimate). New hires follow the guide themselves while HR and Admin check documents and notifications.
- Who can learn from this
- HR, administration, new hires and colleagues designing training or self-service processes.
- Before
- Verbal explanations and lengthy SOPs, with staff guiding setup step by step.
- After
- New hires follow the web guide; HR and Admin verify documents and completion notifications.
Try this approach: make completion conditions explicit so users and the people taking over can both verify them.
Onboarding App link provided; actual access depends on system permissions.
04Personal connections and queries verified
Ask about meeting details without searching transcripts yourself
Team 4 | Tina (Information Security) × Fang (Sales)
Comparable queries take about 3 minutes instead of 5–20 (team estimate). The Agent finds meetings and organizes answers, using each person’s own connection and permissions.
- Who can learn from this
- Sales colleagues, project owners and anyone who frequently checks meeting summaries, action items or project updates.
- Before
- Find relevant meetings, download 1–3 transcripts, then paste them into AI.
- After
- Ask directly; the Agent retrieves transcripts and supplies an answer with sources for human review.
Reusable approach: share the Agent while keeping each user’s identity and data permissions through personal connections.
Practical use and connection scenarios have been verified. Setup guidance is available in Chinese, English and Thai; access depends on individual account permissions.
05Features integrated
Find past tickets faster and reuse handling experience
Team 5 | Purple (Engineering) × Criss (Engineering)
Historical ticket search fell from about 10 minutes to 3 minutes (usage estimate). Search and actions are combined to reduce copying, with records in a consistent structure.
- Who can learn from this
- Engineering, operations and colleagues who often search past cases or prepare handling records.
- Before
- Find information across tools, edit AI responses and copy them back into the ticket system.
- After
- Bring historical search, analysis and record preparation together, with people confirming important actions.
Try this approach: connect repetitive search and preparation steps while retaining result checks for important actions.
Story available; no tool-access link is provided on this page yet.
06Individual use · Wider rollout being tested
A one-page ticket overview shortens morning-meeting preparation
Team 6 | Wade Lee (Tech) × Jane Hu (Tech)
Morning preparation fell from about 30 minutes to 5–10 minutes (lead's estimate). The overview highlights where help is needed, and consistent records support review and handover. Wider rollout remains in testing.
- Who can learn from this
- Team leads, managers and colleagues who consolidate multiple work updates before meetings.
- Before
- Review tickets, emails and handling records one by one before the morning meeting.
- After
- Start with the ticket overview and AI highlights, then confirm where help is needed.
Try this approach: consistently record the situation, progress and next step to support summaries and handover.
Currently used mainly by one lead; not yet formally rolled out to other leads.
02In progress
Cross-department project tracking and progress summaries
Team 2 | Terry (SALES) × Rebecca (ESS)
The team aims to consolidate cross-department progress. Access to some collaboration-platform information is confirmed, while customer identification and data integration are still being validated.
Validation continues: expand usable data sources and confirm customer mapping and update results.
07In progress
Routine maintenance report automation
Team 7 | Johnny (Engineering) × Willie (Engineering)
The team is automating routine maintenance reports and testing receipt completeness when reports from multiple devices arrive simultaneously.
Validation continues: adjust email processing and confirm all reports are received under high volume.
Teams 1, 3, 4, 5 and 6 have shared project stories; other teams continue updating their work. Time figures in these stories are usage estimates, not formal performance statistics. Long-term outcomes remain under observation.
