Back to home

Terms of Service

Service Description

ExtrAct is a document processing service that extracts structured billing data from medical coding documents (HCPCS, CPT, ICD) and delivers the output as downloadable Excel files. ExtrAct processes documents transiently in memory and does not retain uploaded files or extracted patient data beyond the active session.

Zero Retention Architecture

ExtrAct processes your documents transiently and retains nothing. Files enter encrypted, are parsed in memory, and are destroyed the moment your Excel export is ready. We do not train models on your data. We do not keep backups of your documents. We maintain a zero retention architecture that minimizes your regulatory exposure regardless of jurisdiction.

Prohibited Uses

You agree not to:

  • Use automated or programmatic access outside of any officially provided API
  • Conduct competitive benchmarking using the service
  • Reverse engineer the extraction algorithms or prompts
  • Redistribute extraction outputs as a service to third parties
  • Upload documents you do not have lawful authority to process

Account and Billing

Paid subscriptions are managed through our payment partner. You may cancel, pause, or update your payment method at any time through the billing portal accessible from your dashboard. Overages are billed at the rate specified by your plan tier at the close of each billing cycle.

Limitation of Liability

ExtrAct is provided as is with no warranty of any kind. NOMOI Technologies is not a business associate, data processor, or covered entity under any healthcare or data protection framework. You are responsible for verifying extraction accuracy before submission to any billing or clinical system.

Behavioral Data Collection

We collect anonymized behavioral data about how you use the product, specifically for improving extraction accuracy. This includes events like re-running extractions, editing preview cells before download, download timing, and session abandonment. We never record the content of your documents, what you edited, extracted codes, filenames, or anything that could identify patients. All data is stored on our own infrastructure. We do not use third party tracking tools like FullStory, Hotjar, Mixpanel, or Segment.

Automated Product Improvement

Aggregated anonymous signal data is used to automatically refine our extraction instructions over time. The underlying AI model is not trained on your data. Only the instructions given to the model are refined, based on aggregate patterns across all users.

Contact

Questions about these terms: assel@nomoi.ai