Case study · JustEZ · Insurance field inspections · 5 min read
My role
Product designer, sole designer on the JustEZ app
Team
1 designer · 2 business analysts · 4 mobile developers · 2 QA
Timeline & platform
4 months · App for field adjusters
WHAT SHIPPED
photos captioned by AI in about three minutes
disconnected tools folded into one journey
SD cards, folders or manual uploads
from first workflow map to release
JustEZ on iPad: plan the day, capture the evidence, build the estimate
The short version
The problem
Adjusters juggled a claims system, calendar, maps, camera roll, notes and report templates to finish one inspection. The work wasn’t hard. The handoffs between tools were.
What I did
Mapped the end-to-end field journey, set four design principles with the team, and designed the iPad app around five steps: receive, plan, inspect, document and report.
The outcome
One connected workflow. AI captions 80+ photos in about three minutes, media goes straight to the cloud, and adjusters reported completing more inspections per day.
The problem
Before · Disconnected
Claims system
Calendar
Maps
Camera roll
Notes
Report templates
Re-keying the same claim details, moving photos by hand and rebuilding notes into a report every evening.
After · Connected
01
Receive
02
Plan
03
Inspect
04
Document
05
Report
Every step hands its data to the next, so nothing is typed twice and the report is mostly written by the time the inspection ends.
Business objective
Who it’s for
Primary
Field adjuster
Fewer handoffs and less typing, so more of the day goes to inspecting and less to admin. Works one-handed, on site, often offline.
Secondary
Operations
Live job status and complete, consistent claim files that need less chasing and fewer follow-up visits.
Outcome
The business
More inspections per adjuster and faster turnaround from assignment to report.
Approach
01
Connect the workflow
Reduce handoffs between tools and never ask for the same information twice.
02
Remove repetitive work
Use automation only where it meaningfully cuts administrative effort.
03
Design for the field
Clear actions for people who are moving, interrupted and short on time.
04
Keep humans in control
AI suggestions are always reviewable. The adjuster stays accountable.
Key design decisions
01 · Plan
Scheduling weighs priority, location, driving time, inspection length and existing commitments, so adjusters plan a realistic day instead of working down a list of addresses.
Today’s schedule with jobs, route and status in one view
Rescheduling a visit with the impact on the day shown up front
02 · Capture
Photos and video go straight to the claim in the cloud. No SD cards, folders or end-of-day uploads, and a connected external camera records into the same file.
Media organised by room as it’s captured
Recording from a connected external camera
03 · Document
AI suggests photo captions, titles and the inspection overview, and the adjuster reviews every suggestion before it’s saved. Checklists show saved, syncing and incomplete states, so working offline is part of the normal flow, not an error.
AI-suggested overview, edited and approved by the adjuster
Checklists with clear progress and sync states
04 · Estimate
Room dimensions and line items are captured on site, so the estimate is assembled from the inspection instead of being retyped back at the desk.
Room dimensions captured during the walkthrough
Editing a line item without leaving the inspection
Product walkthrough on iPad
Leading across functions
With product
Translated business outcomes into prioritised experience bets for each release.
With engineering
Explored constraints early and specified behaviour, not just pixels, for sync, offline and capture.
With QA
Defined the states and edge cases that protect an adjuster’s confidence in the field.
Outcome
photos captioned in about three minutes, from the public product benchmark
disconnected tools replaced by one journey from assignment to report
scoping reports published to the cloud while travelling between jobs
Takeaway
The two questions I asked of every screen
