Nalyst Research

Two open-source libraries for applied analytics. Nalyst covers machine learning, statistics, time series, survival, and deep learning. Neurova focuses on computer vision and GPU-accelerated imaging. BBox Studio is our local-first, in-browser annotation workspace for computer vision teams—no backend required.

pip install nalyst
pip install neurova

Products

Nalyst: Full-stack ML

One API for supervised, unsupervised, time series, survival, and deep learning with AutoML, explainability, and diagnostics.

  • Unified train/infer pipeline
  • Leakage-aware validation and backtesting
  • Preprocessing, pipelines, and serialization ready for prod

Neurova: Vision & DL

Image processing, feature detection, and deep neural nets with optional GPU acceleration and minimal dependencies.

  • Image I/O, filters, segmentation, detection, tracking
  • Neural layers, autograd, augmentation utilities
  • CuPy acceleration for massive speedups (optional)

Documentation & Resources

Contact

Hemant Thapa
Hemant Thapa
Founder, Nalyst Research (@ Analytics with Harry)

Questions, feedback, or collaboration requests? Reach out anytime.

Based in Glasgow, Scotland.