Artificial

Artificial is a Series A Enterprise SaaS company that helps top-performing brokers and carriers transform their speciality placement and underwriting strategies.

I led the design and frontend for a new ‘Policy Inbox’ to help carriers organise and find important policy information. The project lasted eight weeks, from initial discovery to shipping a beta.

Insurance workers face an uphill battle managing complex policy information. Each policy contains dozens of data points. From coverage details and premium calculations to client communications and renewal dates.

As portfolios grow, underwriters and brokers find themselves drowning in spreadsheets and disparate systems, making it nearly impossible to maintain a clear overview of their work. Policies are moved in and out of their office like an email inbox.

The Inbox is a listing page for all policies, organised by when they were last updated. As the product grew, this caused problems:

• Customers struggled to navigate and find the information they needed
• High-value prospects raised issues about managing renewal policies. This blocked sales.
• Workflows such as Tasks weren’t integrated into the Inbox, making it hard to track work done
• Customers requested custom dashboard due to lack of filtering options
• Lots of bugs and UX issues reported

This led to forming a project team of myself, VP of Product, and Lead Engineer to re-imagine the Inbox from the ground up.

Our goal was to address the above issues, whilst fixing any remaining bugs that stuck around.

I worked with Rob twice as a contractor. He is super skilled, has great taste and brings a ton of experience right across the design ↔ frontend stack.

This guy ships!

Jon Lay, VP Product

What we built

The new Inbox centred around helping customers find information when they needed it. Each policy had over 100 data points, but customers rarely needed all of them at once. Different workflows called for prioritising different information. For example, brokers often wanted to see when policies were up for renewal and who the key contact was.

The new experience focused on:

• Scalable filters that enabled customers to surface the information they wanted
• Saved views that stored those filters in a menu or via a shareable link
• Tasks linked to each policy, with a window to manage and triage them
• Performance improvements through client-side caching so the inbox loaded rapidly

Filtering

Filters are powerful when combined together. This allowed users to create hundreds of different combinations depending on what question they needed to answer. For example, if your work focuses on a particular set of clients, you can group them together and filter out the rest.

As there are often 100+ columns, each representing a data point, we enabled users to customise the order and visibility, omitting or shifting columns based on their workflow.

This allows users to create their own views which they can save for later.

Views

At the top of the page, users can create and manage custom views. Organisations can set up views for their underwriters, and users can create their own.

Each view is a collection of filters and display options applied every time they select it. When you select a view and return to the page, you see the last one you were viewing.

Tasks

Each policy can have tasks associated with it. You can view and create tasks directly from the Inbox, helping you manage your workload faster.

Each policy row shows a badge with the task count. Clicking it opens the task list. If there is only one task, it opens that task directly in edit state. Edits to the task, such as changing the assignee or status, are saved automatically after a short debounce.