04Regulated workflow designHudson Insurance Group, contract · Sep 2025 – Jul 2026
Hudson Insurance
Translating agricultural insurance operations into a workspace where finding an insured and working on one happen at once.
- Problem
- Underwriting for agricultural and livestock insurance means moving between many insureds and their policies. By my account, the earlier model repeated one loop for every record: a table, a detail page, back to the table.
- My role
- Product designer (contract). I joined a design effort already in progress, translated underwriting, regulatory, business, and technical requirements into buildable workflows, and established the product’s interaction foundation.
- Key decision
- Keep discovery and the workspace on screen together: the list of insureds stays in place, and choosing another insured refreshes the workspace without navigating away.
- Outcome
- Stakeholders understood the persistent model immediately, and discussion moved on to what the workspace should show. Shipping status is not documented.
, enlarge to full resolution01
Agricultural underwriting
I joined a design effort already in progress, on a new digital platform for agricultural and livestock insurance.
The work was translating underwriting, regulatory, business, and technical requirements into clear, buildable workflows. The application was built in Angular on an Ant Design implementation I considered dated; I recommended PrimeNG in its place, and the work moved toward a reusable component kit rather than one-off screens.
02
The round trip
The earlier interaction model repeated one loop for every record: a traditional table, a detail page, back, and the table again.
That repeated navigation became the design concern.
, enlarge to full resolution- Table
- Detail page
- Back
- Table
03
Discovery and work at once
In place of the round trip, a persistent discovery panel and a persistent workspace emerged.
Discovery stays on screen. The workspace stays on screen. Selecting another insured refreshes the workspace’s contents in place, and nobody navigates away from either region.
Three decisions that shaped it
Discovery and work happen at once, not in turn
BecauseGoing from a table to a detail page and back again, for every insured, was the repeated navigation to remove.
Trade-offA persistent discovery panel takes about 300px, roughly a sixth of the screen, away from the workspace at all times.
Selecting refreshes the workspace in place
BecauseThe underwriter should keep their place in the list while the content beside it changes.
Trade-offThe workspace has to hold everything about one insured without a page change: contact, actions, activity, policies, coverage.
Build from reusable components
BecauseThe work was moving toward reusable components rather than one-off screens, and I recommended PrimeNG in place of the dated Ant Design implementation.
Trade-offScreens give up one-off layouts in exchange for a shared kit.
1Nothing selected
, enlarge to full resolutionReconstructionBefore a selection: the list of insureds beside an empty workspace. 2An insured selected: the workspace refreshes in place
, enlarge to full resolutionReconstructionThe same list, with one insured’s overview beside it.
In turn
FindWorkFindWork
At once
FindWork
04
An interaction foundation
I established the product’s interaction foundation, creating reusable patterns that simplified complex underwriting and policy workflows.
The design file’s layers name the shared component set the screens were built from:
- Tags and badges
- Data tables
- Panels
- Breadcrumbs
- Tabs
- Timeline events
The design frames carry the names of USDA livestock and dairy programs (LRP, DRP, LGM); the overview’s policies table lists one insured’s policies across them.
05
Feedback
Stakeholders understood the persistent discovery and workspace concept immediately on seeing it.
After that, discussion moved on from the interaction model to the separate question of what content the workspace should show.
Try the interactive reconstructionRebuilt from the two design frames.