My resume

Making property data approachable at Archipelago

Improving access to social services at Healthify

Personal projects

Archipelago

Launching new reports
Product design, data visualization, design systems

The challenge: Archipelago is a platform that centralizes data for insuring trillions of dollars of property. With analytical AI, there’s no reason to depend on spreadsheets.

Even though we already have few metrics and charts in the product, users cannot answer questions about data quality, risk estimates, and costs without spreadsheets or help from our sales and account management teams.

The solution: With reports, users can answer those questions themselves and better prepare themselves in pricing negotiations.

My role: I led the design for all reports, related features, and a new data visualization design system. For each report, I met with internal stakeholders and customers to understand their needs and to validate design decisions.

Impact: Our first launched report sped up deals by 3-4 days per account! Following reports not only saved users hours of data analysis, but also led to several multimillion dollar deals with international customers.

You have questions? We have answers.

As a whole, reports cover a wide range of topics (data quality, pricing estimates, risk factors, and more) and granularity (from portfolio summaries down to individual properties).

Say goodbye to spreadsheets!

Our first one, the Readiness Report, provides property owners and brokers with detailed feedback on the level of data quality that insurers expect. It exposes where there’s missing or poor data and recommends properties to focus on.

Before, the sales team would spend a few days doing diagnostic analyses and creating custom presentations for prospects. Now, we can generate the same content with a few clicks, saving a few days per account.

Figuring out where reports live in the product.

The report at launch. We simplified grading to 3 ranks.

Readiness Report after 2 years of improvements.

We originally gave portfolios a score out of 100 and percentile benchmarking.

We had to take some shortcuts to launch the report. Since this is a template, we’d lose more nuanced customer-specific insights. It’s also meant to be shared as a link or as a 1-2 page PDF, rather than presented as a slide deck.

The Readiness Report was designed to be in a set of 3, but we scrapped the others due to time constraints.

This was originally part of a set.

The cut Engagement Report tracks who looked at your portfolio, when, and which properties.

The cut Enrichment Report shows improvements in data quality.

Our design system now supports data viz.

Since I designed the initial set of reports to go together, I had some semblance of a design system. As we added more reports, I developed chart-specific guidelines for product, design, and engineering to reference. This made it easy to add new chart types, update styling rules, and create high fidelity mockups.

Accessible color palettes for various charts and datasets.

Chart types.

Components and interactions.

Loading and error states.

We introduced new reports over the next 2 years.

I took an iterative approach for each report. When there’s new datasets that our existing charts don’t support, I introduced new chart types. I also tried different layouts based on what could be improved from before.

Risk Assessment Report

We expanded our audience to insurers, who are focused on estimating risk factors. Previously, they would spend ~30 minutes within the product to find the figures they’d need, but now it takes a few clicks to get the same information.

Early designs for a summary dashboard inspired the report.

Tab 1: Geography and hazards.

Tab 2: Construction and occupancy summaries.

Tab 3: Loss history, a new dataset.

Charts are grouped into tabs by topic, so that users aren’t overwhelmed by the amount of information shown. We also introduced new kinds of charts and datasets.

Insurer Transparency Report

We partnered with an international insurer to create this report. They’d supply their proprietary risk grading metrics and bring their ~1600 customers to Archipelago, starting with a pilot of 14 accounts. On our end, we’d process their property data and display their risk metrics on Archipelago.

Early designs from before I joined. The idea was to bridge insurer and Archipelago data.

Here, I organized the metrics by hierarchy through tabs and stacked columns to get users from major categories to individual property data.

Tab 1: Overall grades.

Tab 2: Look into specific risk factors and properties.

Tab 2: Losses were broken down by region and year, instead by property.

Tab 3: Loss history by year and incident.

Exposure Index

This report proactively predicts risk based on exposure to climate and environmental hazards. More ambitiously, we could predict insurance costs based on hazards, losses, and peer comparisons in the future.

The long-term vision: we can predict prospects’ risk levels and where prices should be.

Once they’re customers, we can provide a more accurate report with targeted recommendations.

The shipped version. By default, we display 4 key hazards.

Expand for all hazards and detailed comparisons.

To make the hazard categories easy to scan, I organized information into a table with embedded graphics.

Valuation Summary

Users can check if their cost estimates are close to market values and investigate red flag properties. For this report, I paired with a second designer who wanted to learn about data visualization.

We shared an early design with customers months before we built the integration.

This report goes top down from portfolio sums to
individual properties.

Feedback from recorded customer calls.

From interviews with brokers, we learned that the integration and report are massive time savers. A broker could save at least 6-8 hours preparing a single account for the insurance market. After launch, we saw similar productivity gains among customers.

Say hello to spreadsheets (again)!

Based on feedback from account managers and customers, we added new formatted spreadsheets.

International customers benefited the most from these additions. Since they divide their property management by region and line of business, they need spreadsheets to neatly combine into one view for global risk managers.

Change Analysis Export

Tab 1: Instructions and glossary.

Tab 2: Before vs after comparison.

Tab 3: See every edit per property.

This helps users monitor data edits at a large scale. It shows data changes between two points in time, especially when properties are added, disposed, edited, split apart into multiple line items, or combined.

Readiness Export