Owned by your organization
- Uploaded datasets
- Feature definitions you provide
- Customer-specific models, subject to contract
- Prediction inputs and outputs
- Customer-visible notes and comments
Your data reflects years of decisions and outcomes. The models built from it carry that same value.
Your organization retains ownership of both, along with the predictions they produce. AscentML receives only the rights needed to deliver the service.
AscentML separates what belongs to your organization from the operational information we need to run the service.
This information is used to operate, support, and secure the service.
Without a separate written agreement, AscentML does not:
The advantage in your history should stay with your organization.
We expect to sign appropriate nondisclosure and confidentiality agreements before receiving sensitive business information, subject to normal review of the terms.
Those agreements can define what information is shared, how it may be used, who may access it, and what happens when the engagement ends.
Platform controls support those commitments. They do not replace them.
Customers see what they need to evaluate and use a model: status, performance, dataset lineage, explanations, intended use, limitations, and prediction history.
Hyperparameters, internal experiments, and training internals stay within AscentML’s managed environment by default. That keeps the customer experience focused and protects the methods we apply across engagements—without limiting your ownership of the resulting model.
Data is retained for the life of the contract unless you ask for it to be deleted.
At contract termination, or at your verified request, data can be permanently purged. Purges cover database records, stored files, model artifacts, cached explanations, batch files, and derived exports. Each purge is authorized, audited, and irreversible once complete.
We are glad to start with an NDA. Tell us what your legal team requires, and we will work through it before any data changes hands.