Please wait...
Tarka combines a production-grade ML pipeline, neuroimaging processing, and enterprise collaboration into a single unified platform — so your team can focus on discovery, not DevOps.
From data preprocessing and QC all the way through differential expression, feature selection, and model evaluation — fully automated and reproducible.
WebSocket-powered live progress updates keep your team informed about every pipeline stage as it runs.
Scalable async job coordination handles multiple simultaneous pipelines without blocking your workflow.
Preprocessing, normalization, batch correction, and differential expression analysis on high-dimensional expression matrices.
Native NIfTI (.nii/.nii.gz) support, ComBat and NeuroHarmonize batch correction, and atlas-based regional extraction.
Boruta feature selection, HDBSCAN clustering, LightGBM modeling, and statistical testing for robust biomarker identification.
Plotly and D3-powered interactive charts at every pipeline stage — QC distributions, volcano plots, feature maps, confusion matrices.
Control which visualization types are available per plan or workspace — six policy keys covering all pipeline phases.
Large data files and result artifacts stored in Azure Blob Storage with support for uploads up to 10 GB.
Organize research by teams and projects. Invite members, assign roles, and share results across your workspace.
SSO via Google and Microsoft, multi-factor authentication, policy-based access control, and full audit trails.
Six-level policy hierarchy (app → plan → org → team → user) with numeric limits, booleans, and option lists.
Each stage is an independent microservice — scalable, resumable, and auditable.
Start your free trial and run your first pipeline in minutes.
Get Started Free