Skip to content

ScaiVision: our software platform

At the core of our company is Scailyte’s best-in-class analytical platform ScaiVision™. It unravels hidden secrets of complex single-cell multiomics data to extract composite biomarkers associated with different cell populations. Using a convolutional neural network, ScaiVision automatically learns molecular patterns associated with relevant clinical outcomes. These signatures can then be applied to classify new samples.

ScaiVision Analysis Workflow


We help drug developers increase the success rate of their products and speed up their R&D cycle by:

  • analysing the MoA of drug candidates on a single-cell and multiomics level for mechanistic understanding or to understand differential mechanistic properties vs other drugs
  • identifying predictive markers of response for patient stratification and smarter clinical trial design
  • finding predictive markers of toxicity for improved patient management
  • identifying predictive biomarkers to optimize the selection and enrichment of cell therapy products for optimal response strength and durability and to adequately mitigate product toxicity
  • co-developing a companion diagnostic assay

Key Benefits of ScaiVision

  • Entirely agnostic to indications of interest
  • Data augmentation enables the identification of accurate signatures from a limited number of samples
  • Fast and scalable analysis of datasets of up to hundreds of millions of cells
  • Retains single-cell resolution throughout the interpretation stage & calculates the clinical endpoint-associated score for every single cell
  • Flexible integration of multiomics and clinical data:
    • scRNA-seq
    • CyTOF
    • CITE-seq
    • TCR / BCR-seq
    • High-dimensional flow cytometry
    • bulk RNA-seq
    • bulk proteomics
    • seromics
    • DNA-seq (WES, WGS)
    • ELIspot
    • clinical data

Scailyte’s AI extracts cellular and molecular insights through representation learning

Conventional/non-AI approach:
biased insights & blind spots
Scailyte’s AI approach: 
actionable insights through representation learning
Conventional/non-AI approach: biased insights & blind spots
Scailyte’s AI approach: actionable insights through representation learning

Our supervised representation learning method identifies patterns and classifies cells without prior information on cell clusters and cell types, and can unveil biological signals from previously unknown or rare cells that are associated with given endpoints (e.g. response to treatment).

ScaiVision® performs best-in-class at sample class prediction

Helping you speed up your therapy development and increase your success rate