Skip to content
RK

Web app · Amazon global real estate, commercial portfolio · 2023 to 2026

Jiminy Ticket

A web app that scored a facilities team's ticket data against the client's goals and told each person what to fix.

Confidential client tool, not public.

My role

Designed and built it, ran it 2023 to 2026, handed over the code.

Proof
73 → 2 hrs
first response
36 → 95%
routed right
Skills shown
  • Website management
  • Analytics and reporting
  • Communications project management

The problem

53M+ sq ft · 500+ buildings · worldwide

Amazon's facilities team couldn't see the work behind its tickets. A complex project and a simple fix each counted as one ticket. Routing and priority were inconsistent, and no one could estimate effort, staffing, or why one region ran slower than another.

What I built

A web app where a manager uploads raw ticket data and sees team performance scored against the client's goals. Each person gets the specific fixes that bring their score up to target.

Raw ticket data
Client goals as metrics
Team and individual scores
Each person's next actions
Raw ticket data
Client goals as metrics
Team and individual scores
Each person's next actions

What it measured

  1. End-to-end resolution time
  2. Aging: tickets past a set age
  3. Level of effort, separating complex work from simple work
  4. Simpler routing and priority by category, type, and item
  5. Staffing estimates from volume, type, and priority
  6. Automation, starting with 24-hour intake and simple workflows, then complex ones
  7. Regional differences in completion time

The call

Data showed a lightbulb took far longer to get in Madrid than in San Francisco. Instead of accepting the gap, we set regional SLAs and pre-purchased stock so Madrid could match completion times in the Americas.

Results

First response to new tickets
73 hrs → 2 hrs
First response to new tickets
Routed correctly the first time
36% → 95%
Routed correctly the first time
Tickets open longer than 30 days
−92%
Tickets open longer than 30 days
Average time to resolve
−32%
Average time to resolve
Work capacity, same team
+26%
Work capacity, same team

Approximate figures over 8 months. Exact data is confidential.

The work shifted from routine tasks to complex, strategic ones. I used that shift to make the case for promoting roles to higher levels.

Teams came to us earlier and were more satisfied with the results.

Ran from 2023 to 2026. Amazon retired it for an internal tool built on its model, and I handed over the code.