GlobalData Healthcare is the main division of GlobalData, supplying market intelligence for pharma, biotech, and medical device companies. The platform is built on 290+ proprietary datasets covering drugs, clinical trials, patents and markets. They brought us in to build the frontend. We turned a complex Figma layout into a working web application and built the most demanding charts on the platform.

The platform was built to cut out manual research and help pharma teams make quick, data-backed decisions across four key areas:

• Enabling side-by-side drug comparisons with hyper-specific filtering across dozens of parameters.

• Generating ready-to-use PowerPoint decks for investor meetings and internal updates in just a few clicks.

• Streamlining congress management through an Outlook-synced calendar paired with an AI assistant for presentation insights.

• Visualizing drug pipelines and market trends using interactive Bull’s Eye charts, heatmaps, and timelines.

1. Proven experience with complex data interfaces

We had already built dashboards with heavy filtering and custom timelines, so there was zero learning curve when it came to pharma analytics.

2. An algorithmic approach to visualization

When off-the-shelf charting libraries couldn't handle dense element placement, we wrote custom positioning algorithms. This resolved data overlap issues on both the Bull’s Eye charts and timelines.

3. Tailoring Shadcn UI to exact Figma specs

Instead of simplifying layouts to fit stock components, we extended base Shadcn UI and Radix primitives to achieve full design accuracy without sacrificing page performance.

4. Direct partnership with the product team

Rather than waiting for detailed specs, we brainstormed solutions directly with the client and designers to resolve UI dead ends. We also stepped in to help neighboring teams whenever they hit technical walls.

1. Data collisions on the trial forecast timeline

On the clinical trial forecast timeline, dense data points kept stacking on top of each other, making the chart unusable.

Solution

We ran joint brainstorms with the client’s design and product teams to rethink the track layout logic, landing on a solution that balanced design specs with real-world data density.

2. Overlapping data on the Bull’s Eye chart

High data density caused chart nodes to overlap, and off-the-shelf libraries lacked collision-handling logic.

Solution

We wrote a custom positioning algorithm that dynamically calculates distances between nodes and distributes them evenly across orbits.

3. Calendar performance with 1,000+ events

The congress calendar choked when filtering and rendering hundreds of simultaneous events.

Solution

We optimized state management, implemented list virtualization, and restricted re-renders strictly to visible blocks.

4. Bending Shadcn UI to match a highly detailed Figma layout

Out of the box, base Shadcn UI and Radix primitives didn't support the dense layout requirements from the design team.

Solution

We extended the underlying primitives with custom wrappers, achieving exact design accuracy without adding bundle bloat.

5. Edge-case bugs in React 19 and Next.js 16

Adopting bleeding-edge framework versions boosted performance but triggered compatibility issues with third-party packages.

Solution

We isolated problematic dependencies, patch-fixed incompatible custom hooks, and stabilized the build without reverting framework versions.

Framework: Next.js 16 (App Router, Turbopack), React 19, TypeScript 5

UI: Shadcn UI, Radix UI, Tailwind CSS 4, next-themes, Lucide React, Tabler Icons

Database and infrastructure: MongoDB, Google Cloud Platform, Google Cloud Storage

Utilities: date-fns, cmdk, sonner, class-variance-authority

Export: pptxgenjs

Development tools: ESLint, Prettier, Husky, Vitest, Playwright, SVGR

In six months, the platform went live with a full set of analytics tools for pharma teams.

• 17 unique page routes across 6 major feature modules.

• Bull's Eye and forecast timelines (the two hardest visualizations in the product) delivered exactly to design, including the one another team couldn't ship.

• A custom component layer on top of Shadcn that holds a dense design without bloating the codebase.

• The frontend delivered on schedule with minimal revisions, working in step with the client's design team.