
airhacks.fm podcast with adam bien Building Software for Chemistry Labs with Java
Dec 10, 2025
Stanislav Bashkyrtsev, a software developer and founder specializing in scientific software for chemistry and drug discovery, shares his insights on the challenges of automating lab processes. He discusses Pixel, a powerful tool for analyzing mass spectrometer data, and the intricacies of parsing large instrument files. Topics include the drug discovery workflow, molecular library enumeration, and the significance of 2D vs 3D visualizations. Stanislav also introduces innovative solutions like Molbrett for chemical drawing and Crystalline for protein crystallography documentation.
AI Snips
Chapters
Transcript
Episode notes
Mass Spectrometry Decoded
- Stanislav explains mass spectrometer data reveals molecule masses by flight time and peak patterns.
- His software parses vendor binary files and predicts molecules from isotope patterns.
Prepare For Huge Measurement Files
- Analytical files vary hugely: megabytes to gigabytes per run.
- Efficient parsing and storage design are essential for performance.
Design Storage For Scale
- Start with Postgres for simplicity but plan for scale limits.
- Use object storage (S3) for large raw files and handle concurrent uploads.

