Brookhaven’s Tiled Tool Simplifies Scientific Data Access for Researchers

Last updated: October 15, 2026

Scientists at Brookhaven National Laboratory's NSLS-II have released Tiled, an open-source data access service that lets researchers browse, preview, and extract specific data without downloading entire datasets. The tool addresses a persistent pain point: large scientific experiments generate enormous data volumes, and existing methods force researchers to download everything before determining what matters.

Tiled sits between databases, file systems, and user applications, providing a consistent interface regardless of where data lives.

At a Glance

• Tiled is a data access service for scientific portals and data science tools
• Developed at Brookhaven's NSLS-II, which hosts nearly 2,000 scientists annually
• Supports multiple data formats, sizes, and access patterns
• Published under an open-source license for use at labs worldwide
• Built in Python but accessible via API from any language

How Tiled Works

NSLS-II welcomes scientists who probe samples with X-rays and capture data through advanced detectors. Previously, PhD advisors helped download all data at once before classifying useful portions—tedious work when each detector type requires different transformations.

Tiled lets researchers preview data in a browser or analysis application, select only what they need, and download in their preferred format. They can also feed data directly into analysis software without manual file handling.

The service uses an API-based architecture. Clients connect through defined rules, allowing developers to build tools that extend Tiled's capabilities beyond what the core team originally envisioned.

Technical Design

Tiled integrates naturally with Python-based data science libraries while remaining accessible from other languages. The API approach decouples data storage from data consumption, meaning scientists can access files, databases, or web services through the same interface.

Tiled also relates to Bluesky, data acquisition software developed at the same facility. Both projects aim to make scientific data workflows more accessible and reproducible.

Our Analysis

For developers building data pipelines or research tools, Tiled demonstrates API-first design principles. The service doesn't assume a single data format, storage backend, or client application. Instead, it provides a consistent interface that adapts to existing workflows.

This approach matters for any team building tools that must span different data sources. Tiled's open-source release and Python foundation make it accessible for developers who want to study or extend the architecture.

Practical Example

Say you're building a dashboard that aggregates sprint metrics from multiple sources: Jira for stories, GitHub for commits, and a custom database for retrospective notes. Instead of building custom connectors for each source, create a unified data access layer. Scientists working with Tiled have shown that standardizing the access interface—even if the backends vary—simplifies client development dramatically.

What to Watch Next

Tiled's open-source release means broader adoption is likely. Watch for integration with common scientific Python packages and additional facility deployments beyond Brookhaven.

Key takeaways

• Tiled enables selective data access without full dataset downloads
• API-first design allows client flexibility regardless of programming language
• Open-source licensing supports adoption across research institutions
• The Python foundation integrates well with common data science workflows