Built for Higher Education. Designed for Control.

AI Agents access student data that your institution has already loaded into Mongoose through your existing CRM or SIS integrations – such as student name, contact information, application status, and any custom fields you’ve configured. Agents also use conversation history and institutionally approved knowledge sources (your website content, uploaded documents) to generate responses.

The AI model itself processes this data solely within the context of generating a response – it does not store, learn from, or retain any of this information after the interaction is complete.

No. Mongoose does not use student data, conversation content, or any institutional data to train, fine-tune, or improve AI models – under any circumstances. This is a core commitment that applies across all of our products and use cases, without exception.

Our AI infrastructure runs on Azure OpenAI Service, which enforces zero data retention. Microsoft’s terms explicitly state that prompts, completions, and embeddings are not used to improve OpenAI models or any Microsoft or third-party products.

All data is stored and processed within Mongoose’s private Microsoft Azure environment. AI models are hosted through Azure OpenAI Service within this same environment – we do not use consumer AI services like ChatGPT or any public AI endpoints.

No. Because we use the Azure OpenAI Service, hosted within our own Azure environment, student and institutional data never leave our infrastructure for AI processing. Azure OpenAI enforces zero data retention – prompts and outputs are not stored beyond the immediate API call. No data is shared with OpenAI, Microsoft, or any third party for model training, improvement, or any other purpose.

Student data is protected through multiple layers:

• Prompt-only usage: Student data is passed to the AI model solely as context to generate a response. Once the response is generated, the data is not retained by the model or the model provider.
• Zero retention: Azure OpenAI does not store prompts or completions beyond the immediate API call.
• No public model exposure: Data is never sent to public or consumer AI services.
• Input/output filtering: All inputs are screened for malicious patterns before reaching the model, and all outputs are validated against defined schemas and content filters before delivery.

Yes, all AI-driven conversations are logged and available for authorized staff to review within the Mongoose platform. Conversation data is retained for the duration of your contract or at your institution’s direction. You may delete student data, messages, and conversations at any time.

Institutions also have granular control over data – you can delete individual contacts at any time, and associated conversations will be removed accordingly. Upon contract termination, audited data deletion occurs within 30 days. After deletion, only aggregate data (such as message counts or agent activity summaries) may remain for reporting purposes – this never includes student data, messages, or conversation content.