Frequently asked questions

Ounias is designed primarily for researchers and teams working with structured data, including biomedical, life-science, and other research settings.

It can also be useful anywhere the goal is to understand what a dataset contains, compare alternative analytical approaches, and preserve how conclusions were reached.

You do not need to write code or be a machine-learning specialist to use Ounias.

The platform provides profiling, guidance, sensible defaults, and evaluation tools while keeping important analytical choices and evidence visible.

Researchers remain responsible for interpreting results in the context of their data and research question.

Ounias can recommend compatible approaches and provide sensible defaults, but important analytical choices remain visible and, where appropriate, configurable.

Its purpose is to support investigation and comparison rather than hide the analysis behind a single automated result.

Ounias is organized around investigations rather than isolated model training.

Analyses remain connected to their datasets, configurations, lineage, evaluations, comparisons, and derived work so that alternative approaches can be examined as parts of the same investigation.

Ounias can automate parts of an analytical workflow, but its goal is not simply to return a single optimized model.

Ounias can be used to analyze biomedical and clinical research datasets, subject to the data-governance and privacy requirements of the researcher and their organization.

Ounias is a research analysis platform, not a medical device, and is not intended to diagnose, treat, or determine the clinical management of an individual patient.

Ounias is focused on structured, tabular datasets. Data can include continuous, categorical, binary, supervised-target, and longitudinal information.

Supported import options currently include common tabular file formats and connected sources such as Google Drive.

Your data remains yours. Ounias does not claim ownership of datasets you upload or customer-specific analytical outputs generated from them.

You can export your datasets and analytical outputs or delete them when you no longer want them stored in Ounias.

No. Ounias may fit or train models when you request that as part of your analysis, but your datasets are not used to train general Ounias models, models for other customers, or for unrelated internal research.

Ounias does not sell Customer Data, share it with advertisers, or use it for behavioral advertising.

Raw Customer Data is not sent from Ounias to connected third-party services. Where an outbound integration is supported, only the analytical outputs, models, summaries, artifacts, or other results you select are transmitted.

Read more about privacy →

Ounias uses technical and organizational safeguards intended to protect Customer Data and restrict access to authorized users and services.

Access controls, authentication, infrastructure protections, and data-handling rules are designed so that analytical data is available only where it is required for the operation you request.

Additional security and deployment requirements can be discussed for institutional use.

Where possible, yes. Ounias encourages using only as much identifying information as an analysis actually requires.

For example, if age is sufficient for a research question, using age rather than an exact date or year of birth can reduce unnecessary identifying information.

Whether a dataset is sufficiently de-identified depends on its content and context, so that determination remains with the researcher or organization providing the data.

No. Your datasets and analyses are not made public simply by using Ounias.

Public sharing occurs only through features intended for that purpose, such as publishing a Snapshot.

Ounias keeps your work for up to 60 days after a subscription ends so that an accidental lapse does not immediately erase your datasets and analyses.

After that period, Customer Data is deleted in accordance with the Ounias retention policy.

Ounias supports a growing set of supervised and unsupervised workflows, including clustering, dimensionality reduction, supervised learning, explainability, stability assessment, substructure analysis, longitudinal analysis, and comparison across analytical runs.

Available methods and capabilities may depend on the selected plan and the characteristics of the dataset.

Not in the sense of declaring one universally correct answer.

Ounias profiles datasets, recommends compatible approaches, provides evaluation workflows, and makes it easier to compare alternatives across multiple forms of evidence.

A run is a persisted analytical result created from a particular dataset, method, and configuration.

Runs preserve their metrics, artifacts, settings, and lineage, allowing analyses to be reproduced, compared, and traced back to the work that produced them.

Ounias preserves analytical runs together with their dataset relationship, configuration, metrics, generated artifacts, and lineage.

This makes it possible to revisit an analysis, understand how a result was produced, and compare it with subsequent work.

Yes. Comparison is a core part of Ounias. Runs created with different methods, configurations, embeddings, or variations can be inspected side by side using standardized metrics and analytical views.

A Snapshot is a frozen analytical representation of selected work that can be published through a persistent public link.

Snapshots are intended for sharing, reference, auditing, and preserving the state of an analysis without exposing the underlying raw dataset.

Yes. Ounias supports exporting datasets and analytical outputs, and selected workflows can connect results to external tools.

Ounias is currently working with researchers and organizations through early pilots.

Apply for a pilot →

Ounias uses subscription plans with included capabilities, storage, and compute. Additional capacity or standalone products may be available separately.

View current pricing →

Yes. Ounias is designed to support both individual researchers and research teams.

Institutional arrangements can be discussed where additional deployment, privacy, security, or workflow requirements apply.

Contact Ounias →

Organization-specific deployment arrangements may be available where additional infrastructure, privacy, security, or data governance requirements apply.

Discuss deployment requirements →

Ounias is being designed for both individual and team-based research workflows.

Available collaboration capabilities may depend on the plan and deployment arrangement.

Send a message to support@ounias.com or visit the contact page.