Drop in a phone photo. Petrilya finds every colony, gives you the count, and exports a clean report. Open-source desktop app. No code, no cloud, no GPU required.
Three clicks from a photo on your phone to a report on your director's desk.
Drag a JPG/PNG/TIFF from your microscope, phone, or scanner into the window. Or process a whole folder at once with batch mode.
Cellpose under the hood. Pretrained models segment every colony, no settings to tweak. Each one gets an ID, area, diameter, eccentricity.
One-page PDF with preview, histogram, and table. CSV for spreadsheets. JSON manifest with SHA-256 and parameters for reproducibility.
Drop a JPG, PNG, or TIFF straight from your phone or microscope. Batch a whole folder when you have a hundred dishes.
Powered by Cellpose under the hood. Pretrained models recognise colonies of different shapes without any setup.
Erase a misdetected colony with a click. Paint in one the model missed with the brush tool. Metrics recalculate instantly.
Wheel-zoom into any part of the dish, Space-drag to pan. Inspect small details, then fit-to-window with F.
Set micrometers-per-pixel and your areas come out in µm² and diameters in µm — ready for the methods section.
One-page PDF report with preview, histogram and table. CSV for spreadsheets. JSON manifest with SHA-256 for reproducibility.
Windows, macOS, Linux. CUDA when you have it, plain CPU when you don't. Your data never leaves the machine.
Point at a folder, get a CSV, PDF and JSON per image plus a summary.csv aggregating the run. Walk away while it works.
No telemetry, no accounts, no upload. Sensitive patient samples and pharma data stay on the workstation.
Petrilya is in pre-alpha. There's no installer yet — the project is being built in public. To try the in-progress build, clone the repository.
git clone https://github.com/petrilya-app/petrilya-core.git
cd petrilya-core
python -m venv .venv
.\.venv\Scripts\Activate.ps1
pip install -e .
petrilya-uiPetrilya is dual-licensed: AGPL-3.0 for the open-source build, and a commercial license for organisations that need on-premise deployment without copyleft obligations — including air-gapped lab networks and GMP-relevant workflows.
Interested in a pilot, custom integration with your LIMS, or training data? iliarogogvykh@gmail.com