Balthazar¶

Balthazar is an operating system for deep tech R&D labs. Samples are defined once, experiments are written as Python, and every Run keeps its code, Inputs, results, plots, and logs together, so the work stays traceable long after the measurement.

It replaces the usual mix of ad-hoc scripts, folders of data files, and spreadsheets. These docs explain how to use it. They assume knowledge of the lab, not of programming.

The three building blocks¶

• Flows: Python scripts that define an experiment or an analysis.
• Devices: the samples being measured, such as a chip, a wafer, or a cell.
• Runs: one execution of a Flow, with its data, plots, and logs saved, plus its Inputs and Outputs.

The Python code runs on the computer where the Runner is installed, so a Flow can use any library installed there. The Flow itself is controlled from the Balthazar web app, so it can be started from anywhere.

Bringing existing code¶

An existing script pasted into a Flow runs as it is. Two small changes make it work the Balthazar way:

• Read the values that vary between measurements from Inputs (blt.params), so they can be changed before a Run without editing code.
• Save the key results as Outputs (blt.output), so they can be compared across many Runs.

Logs are captured automatically, plots appear once the script calls plt.show(), and the names of any files the script writes are recorded.

Where to start¶

Getting Started covers installing a Runner and connecting it to a Space. From there, Flows and Runs covers writing and starting a first experiment.