Neural Gateway

Me and Your Data

A transparent breakdown of what is tracked, what isn't, and how long it stays.

Last Updated: January 2026

Core Privacy

1. Model Prompts & Responses

The primary purpose of this gateway is to process AI requests. However, I treat the actual content of your conversation as strictly ephemeral.

I do not collect your prompts or the model's responses. Data sent to the API is passed immediately through to the inference engine and returned to you. It is not saved, archived, or used for training.

Account Information

2. What I Do Keep

To maintain the service, manage access, and monitor performance, the following data is stored in the database:

  • Identity: Your email address and a secure hash of your password.
  • Usage Metadata: I track which models you use (e.g., "llama3", "mistral") and the model names.
  • Performance Metrics: The time it took for the model to generate a response (latency/generation time).

This metadata is used solely to monitor system health and debug slow responses.

Telemetry

3. Hardware & System Specs

At my discretion, the client software may collect system information regarding the machine accessing the gateway. This generally includes:

  • Computer specifications (CPU, RAM, GPU)
  • Operating System version
  • Client version

Opt-Out Available: This data collection is not mandatory for service usage. You may opt out of hardware telemetry via your account settings or client configuration.

Data Lifecycle

4. Retention & Deletion

If you choose to delete your account, your personal data is not removed immediately.

  • Retention Period: I retain your email, password hash, and historical usage logs for up to 6 months after account deletion.
  • Purpose: This is to prevent abuse, ban evasion, and to assist in any retroactive security investigations.

After this 6-month period, personally identifiable information is permanently purged from the database.