Designing an insight-driven underwriting experience

The team had earlier developed an underwriting solution, which was clunky and unintuitive. However, we quickly realised the untapped potential in developing a solution that can bolster the user's decision-making process.

We began with the initial vision of developing an MVP, which we converted into a full-fledged system with multiple modules.

Goals

1. Unify data and automate rules to streamline decisions and reduce manual effort.
2. Streamline data consolidation for quicker, precise risk assessment.
3. Increase automation and data accuracy to boost efficiency and speed.

THE PROBLEM

Navigating fragment data landscapes is time-consuming and inefficient

Underwriters struggle with fragmented information across emails, documents, and legacy systems, making it time-consuming to sift through and increasing the risk of missing critical signals.

How can we develop a unified, data-rich workspace that enhances efficiency, provides deeper insights, and facilitates quicker, smarter risk assessments?

Hypothesis

By consolidating all submission data, underwriters can make informed decisions more easily, boosting productivity and confidence.

BUSINESS OPPORTUNITY

Transforming data chaos into an advantage

The team felt that this was our chance to turn data chaos into a competitive edge. By centralizing submission details through intelligent ingestion and layering in third-party data, we give underwriters a clear, unified view. With appetite and referral checks automated, teams can move faster and focus on what matters most, making smarter, high-value decisions.

THE RESEARCH

Uncovering insights

Expert interviews and literature reviews revealed that Underwriters struggle to make informed decisions due to the overwhelming and often inaccurate submission data. These insights refined our understanding of the domain and persona, highlighting their needs in our solution.

Research Insights

Gap Analysis

We conducted a gap analysis on an internal product to align our current state with our target. While we chose to keep some features, we needed to refine the flow and orchestration. The identified gaps guided our focus in developing the MVP.

Success Metrics

Before jumping into ideation, we defined success metrics to align on what impact looks like. These metrics help us track long-term effectiveness, measure efficiency gains, and ensure the solution drives meaningful outcomes for underwriters and the business.

IDEATION

Aligning the end-to-end flow

We focused on high-impact modules essential for the MVP to keep our timelines on track. I created a process flow taking into account our architecture, constraints, and inflow of data at each stage.  We had 2 workshop sessions to identify the key stages and break them down into actionable tasks for the underwriter in each stage. Since documents were never available simultaneously, it was crucial to design the process flow to build confidence in underwriters and ensure they access all critical information for their decisions.

Enhancing Workflow Transparency

We needed a way to inform users about the remaining steps before closing the submission. Underwriters must use a rate checker and input the final quote into a third-party system due to integration limitations.

To ease their concerns and keep them grounded in their decision-making, we incorporated the stepper as a persistent feature throughout our application.

Grid Layout

Our main objective was to streamline decision-making by presenting both primary data (submissions) and secondary data (third-party sources) through an insights dashboard. 

We were initially inspired by the Bento Grid layout. However, after thoughtful evaluation and collaborative design workshops with the engineering team to assess widget implementation, we opted for a structured grid layout. Widgets were standardized into three sizes: Small (S), Medium (M), and Large (L), with clear labeling for consistency.

Fixing Layout with Rigid Widgets

We designed over 40 widgets of various sizes for the insights page, including those sourced from submission documents, verified third-party data, and business-specific widgets. This provides users with a comprehensive view of submissions.

Underwriters required a method to compare submissions across multiple data points to assess risk and understand the business foundation. We developed several concepts for a comparative analytics widget to enhance its functionality and better showcase differences.

While designing the widget to align with its core intent, we faced challenges with data density and layout constraints within the grid. To maintain layout integrity, we introduced a locked mechanism in the insight view management module. Alternatives like fixed height with scroll were explored but proved ineffective for optimal content consumption.

After several iterations and research, we opted to transform the line and dot representation into a bullet graph. This design communicated the selected submission with a filled bar, while the vertical line indicated the comparative option.

This adjustment allowed us to align all widgets with the widget framework, improved visibility of other widgets above the fold, and reduced restrictions on the insight view management module.

Other Features

Inline Rule

The rule engine served as an appetite checker, auto-declining non-compliant submissions. We added inline rule editing within the table, streamlining rule management and boosting efficiency.

View More Data

Underwriters receive multiple documents for each submission. Our submission documents page offers a summarized view of these documents. To enhance their analysis, we designed a screen that allows users to create custom data cuts for clearer insights.

FINAL DESIGNS

Our MVP Solution

We successfully delivered an MVP of the Intelligent Underwriting solution, achieving our goals of automation, information consolidation, and insight presentation.

Submission Summary

Users get an overview of the submission details, track progress using the intuitive stepper, and can take notes to share key milestones with the underwriting team. By leveraging Generative AI, users can view insights on loss history, asset details, and operational data.

Submission Insights

Submission insights, drawn from primary and secondary data, empower users to make informed decisions, evaluate risks, and view all submissions in one consolidated location.

Submission Data

Consolidates documents, organizes data into relevant categories, summarizes information for quick access, and allows for custom data cuts based on specific needs.

Rule Engine

Automates appetite checks, enhancing efficiency and streamlining decision-making by reducing manual effort and applying defined business rules.

View Management

Enables managers to tailor submission insights for underwriters by line of business. This approach diversifies and prioritizes widgets according to submission complexity, ensuring critical informative widgets are prominently displayed.

THE IMPACT

What we achieved

Our solution enhanced underwriters' efficiency and decision-making. Since launch, we've refined the rule engine, created a smart document ingestion module, and leveraged AI to provide valuable insights for underwriters and managers.

~99%

faster quote generation time

60%

reduced cycle time

25%

improved quote-to-bind ratio

40%

reduced overhead costs

01

Future enhancements and client engagement

Building on the MVP's success, we enhanced features, introduced new personas, and pitched our solution to more clients, resulting in increased traction and visibility across the organization.

02

Overcoming challenges with curiosity and teamwork

The complexity of the domain, tight timelines, and lack of real user access posed significant challenges in strategizing the design. However, my curiosity, transparency, and balanced collaboration greatly contributed to delivering impactful results.

03

Research is key to success

This project taught me that research is vital for success. While I prefer secondary research and user interviews, expert reviews are equally valuable, adding another method to my toolkit.