Introduction

Relay is a research workspace that runs in your browser. Your data — images, video, recordings, single-cell matrices, genomes, tables, PDFs — opens in the same place you analyse it, write it up, and share it. There is nothing to install, and the heavy lifting runs on cloud GPUs you pay for by the minute.

It is built for scientists, not for people who set up servers. If you can use a desktop computer, you can use Relay.

Start here

Quickstart

Sign in, open a file, process it, share the result. Ten minutes.

The Workspace

The desktop, the five apps, the menu bar, and Ray.

Data

Where your files live, what you can connect, and what Relay opens.

Ask Ray

The assistant at the bottom of the screen, and what it can do for you.

What Relay does

• Opens scientific files as they are. Microscopy (OME-TIFF, CZI, ND2, LIF, Zarr), DICOM, NWB and SpikeGLX recordings, video, .h5ad and Seurat objects, FASTQ/BAM/VCF, FCS, molecules, PDFs and tables — without converting them first. See Viewers.
• Processes images without code. Two hundred operations — blur, threshold, watershed, FFT, colocalization — run in your browser the moment you click and stack as layers. See Processing tools.
• Runs your code too. Jupyter notebooks in Python or R, with a runtime you switch from CPU to A100 in one click. See Notebooks.
• Turns results into a paper. Slides, manuscripts and posters built from the figures you already have open. See Studio.
• Shares the whole desk. One link sends the file, the zoom, the layers and the annotations exactly as you see them. See Sharing.
• Has an assistant that can drive it. Ray opens files, runs tools, writes notebooks, and answers questions from these docs. It works from the chat bar, Slack, the command line and your own scripts. See Ray.