Web Application
Meridian Labs
Making 47,000 experiments legible.
Meridian Labs
Web Application Development · Data Visualization
Research & Labs
The brief
Meridian Labs had four years of experimental data — 47,000 runs across twelve material families — sitting in CSV exports that no one outside the original researchers could read. New team members took six weeks to find their way around. Industry partners reviewing the research came in for one-day visits and left without the context they needed.
They needed an internal-facing application that turned the archive into a navigable, queryable, visual landscape. Not a dashboard. A tool for thinking.
Approach
We worked alongside three researchers for two weeks before writing a line of code. The interface had to mirror how they already thought.
Cognitive interviews
Shadowed researchers using the existing CSV workflow. Mapped the questions they actually asked of the data versus the ones the schema let them ask.
Data normalisation pipeline
Built an ETL that ingested the historical CSVs, reconciled inconsistent units, and produced a unified queryable store.
Visualization engine
D3-based exploratory interface with linked scatter, parallel coordinates, and a custom material-family treemap. Every view brushable, every filter shareable as a URL.
Onboarding pathway
Built-in tutorial mode that walks a new researcher through three exemplary investigations before they touch the live data.
Results
experiments now navigable
weeks to onboard a researcher
industry partner sessions
patents identified in archive
Stack
Next.jsD3.jsPostgreSQLPython ETLVercel
It changed what our archive is for. Data we had written off as historical context is now actively driving new research directions.
Have a project in mind?
Tell us what you’re building and we’ll come back within 48 hours.
