Understand your data

Ounias is the white-box ML environment designed for:

Enhanced Exploration

Compare models, view projections side by side, and return to any thread of analysis instantly — no retracing needed.

  • Cluster using K-Means, HDBSCAN, Agglomerative, Spectral, GMM, and more coming soon.
  • Project with PCA, t-SNE, UMAP, Autoencoder, DEC, and more.
  • Visualize outputs on interactive 2D or 3D plots.
  • Access run lineage — a visible, organized reasoning trail.
  • Verify results with Silhouette, DBI, CHI, ARI, p-values, and more.

Clutterless Clarity

Evaluate runs, clusters, and observations without losing your footing. Statistics, feature contributions and stability measures share one space, built for focus and interpretation.

  • Compare algorithms and parameters across methods.
  • Analyze clusters — interpret contributions, stability, and statistics side by side.
  • Inspect neighborhoods and local density.
  • Focus on individual observations or patients.
  • Track subjects over time [coming soon].
  • Predict future behavior [coming soon].

Reliable Research

Analyze consistently, even as datasets evolve. Collaborators can verify results without accessing raw data, and snapshots preserve a complete lineage from upload to publication.

  • Update datasets via Google Drive to avoid stale data.
  • Generate reproducible runs every time.
  • Share exact computational lineage.
  • Export datasets, projections, runs, and plots with one click.
  • Publish immutable snapshots [coming soon].

No Strings Attached

Start with the essentials — or enhance your analysis with advanced methods, GPU acceleration, feature contributions and more.
Public beta introductory prices: plans start at $19/mo, cancel anytime.

  • Utilize latent-space learning and advanced clustering algorithms.
  • Increase compute and storage quotas for larger datasets.
  • Unlock feature contributions, stability bootstrapping and more.
See the demo → Explore plans →

An Evolving Environment

Ounias is expanding — new features, more methods, and improved transparency. Here’s what we’re working on:

Supervised Methods

Train custom models with labeled data, alongside the existing unsupervised suite.

Q1 2025

Snapshots

Create an interactive, publication- and audit-ready artifact of your analyses.

Q1 2025

LLM-powered querying

Use metrics and run outputs to analyze datasets using natural language.

Q2 2025

Trajectory Predictions

Track observations over time, enabling trajectories and enhanced predictive modeling.

Q2 2025

*Roadmap is indicative and may evolve with user feedback and emerging research directions.

Do I retain ownership of my data?

Yes. You always retain full ownership of your data and can delete it at any time. Ounias never uses your data for analytics, research, or model training.

Can I use Ounias for academic or clinical research?

Absolutely. The platform is built for transparent, reproducible analysis and is suitable for research workflows, exploratory analytics, and publication pipelines.

Does Ounias store my raw data?

Yes — securely, behind Google Cloud infrastructure. You may also manually delete any dataset at any time.

Can other users see or interfere with my data?

No. Each dataset is processed in an isolated execution environment with no shared memory or temporary storage across users.

Powered by Google Cloud — all datasets are stored securely behind Google’s infrastructure. Users retain full ownership and can delete data at any time. Ounias never uses your data for analytics and does not share it with third parties.

Read the Privacy Policy →