Privacy Protocol.

At Praveg AI, we treat your data with the same rigor we apply to our code. We act strictly as a Data Processor. This Privacy Policy outlines how we collect and process information when you deploy our AI Workforce.

1. Data Processing & The Neural Uplink

We process specific data streams solely to power your autonomous workforce:

• Identity Data: Name, email, and business details provided during account creation.
• Voice Artifacts: Temporary audio streams and transcripts of end-user interactions with the AI Agent.
• Synapse Routing: Real-time API routing of CRM data. We act as a bridge; we do not clone your CRM databases into our servers.

2. TRAI/DoT Compliance & Enterprise Trunks

To ensure telecom network integrity and prevent spam, Praveg AI provisions dedicated +91 Enterprise Trunks for our clients. We must comply with strict government KYC (Know Your Customer) regulations.

• Before a line is activated, the Data Fiduciary (Client) must securely provide: Company Incorporation Certificate, GST/PAN credentials, Authorized Aadhar, and a signed CAF (Customer Application Form).
• These documents are submitted directly to the official telecom providers and are not used by Praveg AI for any marketing purposes.

3. Strict Prohibition on Global Model Training

Your intelligence belongs to you. Praveg AI strictly prohibits the use of your private conversations, sales scripts, or customer interactions to train generalized global AI models.

Brand-Voice Cloning feature available starting from enterprise Plan requires verified legal consent from the company as well as the concerned executive and do not fall under the category of AI training.

Processed data is used only to facilitate real-time contextual awareness for your specific AI Agent during a live call.

4. The 48-Hour Volatile Memory Protocol

Transparency and Data Minimization are absolute. Praveg AI enforces a strict self-destruct mechanism on all call data to protect end-user privacy.

• Automated Destruction: Call audio and transcripts are retained in our secure dashboard for a maximum of 48 hours to allow the Client to audit AI performance. After exactly 48 hours, these artifacts are permanently deleted from our servers.
• Zero-Retention Mode: Enterprise clients may enforce strict "Zero-Retention Mode." In this state, audio is processed entirely in RAM and discarded within milliseconds of the call ending. (Note: Activating this will disable your Sentiment Analytics and Transcript generation).

5. Communication & Grievance Redressal:

For privacy inquiries, security audits, or to exercise your rights , please contact our Intelligence Division at: