Scientist Guide

Scientist Guide

As a scientist or data analyst, you use Brane to compose and run workflows — sequences of operations that process data across distributed domains. You don’t need to know about containers, infrastructure, or networking. You just write what you want to compute, and Brane handles the rest.

What You’ll Learn

1. Writing Workflows — How to write BraneScript workflows
2. BraneScript Reference — Language syntax and features

Prerequisites

• The Brane CLI installed on your machine
• Access to a running Brane instance (ask your administrator for the address)
• Packages already published to the instance (ask your developers)

Getting Started

Connect to an Instance

brane login http://<INSTANCE_ADDRESS> --username <YOUR_NAME>

Explore Available Packages

See what packages are available on the instance:

brane package search

Start the REPL

The interactive REPL lets you run BraneScript commands one at a time:

brane workflow repl --remote http://<INSTANCE_ADDRESS>:50053

Or run a workflow from a file:

brane workflow run --remote http://<INSTANCE_ADDRESS>:50053 workflow.bs

Workflow Basics

A typical workflow:

1. Import the packages you need
2. Call functions to process data
3. Print or store the results

import data_loader;
import ml_trainer;

let data := load_data("patient_records");
let model := train(data);
println(model);

Brane figures out where each step should run based on data location and policies. Your workflow looks like a simple sequential program, but it may execute across multiple organizations.