1. Fundamental Philosophy
Vedic AI processes extremely sensitive data—birth coordinates, personal consultations, and biometric topological scans. We do not treat privacy as a compliance checklist; it is the foundation of our engineering architecture. Our primary business is advanced astrological intelligence, not data brokerage. We do not sell your data, nor do we monetize it through third-party advertising networks.
However, we are a research and product company actively developing foundational artificial intelligence engines specializing in Vedic Sciences. To achieve unparalleled precision, our proprietary models must learn. This policy outlines exactly what we collect, how it is used to train our sovereign AI infrastructure, and how you retain ultimate control.
2. The Information We Collect
We strictly limit data collection to what is necessary for our computational engines to function effectively.
A. Essential Computing Data
- Authentication & Profile: Primary identifiers required for secure account access (Google OAuth IDs, display names).
- Origin Coordinates: Exact date, time (to the minute), and precise geological coordinates (latitude/longitude) of birth necessary for calculating observatory-grade ephemerides.
- App Telemetry: Device identifiers, OS versions, and crash diagnostics to stabilize our mobile infrastructure and compute pipelines.
B. Interaction & Synthesis Data
- Consultation Artifacts: The textual contents of your queries, AI responses, generated Muhurta calculations, and session histories across our various engine nodes (Kundali, Sastri, Guru).
- Biometric Extraction (Image Processing): When utilizing Palm or Face reading algorithms, images are temporarily captured for computer vision analysis.
3. Proprietary AI Model Training & Fine-Tuning
Vedic AI operates highly specialized proprietary language and reasoning models designed specifically for Jyotish. To continually elevate the reasoning capabilities, astrological pattern recognition, and consultation quality of our engines, we utilize user interaction data.
Under our Data Iteration Protocol:
- Model Fine-Tuning: De-identified conversational transcripts, prompts, and the resulting astrological outputs are aggregated into secure datasets. These datasets are used to train, fine-tune, and align our proprietary AI models.
- Anonymization: Before any transcript enters our training corpus, it undergoes an automated scrubbing pipeline designed to strip Personally Identifiable Information (PII) such as specific names, phone numbers, or addresses.
- Your Opt-Out Right: We fundamentally respect user sovereignty. You have the explicit ability to exclude your consultation data from our model training pipelines. This Opt-Out toggle is permanently accessible within the Settings menu of the Vedic AI mobile app. Opting out will not degrade the quality of inference in your active sessions, but your data will no longer contribute to the collective improvement of the Vedic AI architecture.
4. Computer Vision and Asynchronous Image Processing
Advanced features such as Palmistry and Facial structure analysis rely on high-dimensional computer vision processing. To achieve observatory-grade precision, we utilize asynchronous multi-engine synthesis.
- Extraction Window (30-Day Retention): Due to the immense computational requirements of cross-verifying biometric topologies against classical Jyotish parameters, uploaded images are temporarily stored in our secure processing queue. We retain the original source photograph for a strict maximum of 30 days to allow our proprietary AI engines to complete deep topological extraction.
- Source Purge Protocol: Upon the expiration of the 30-day processing window, or immediately upon the successful extraction of necessary geometric vectors (whichever occurs first), the raw source photograph is permanently destroyed from our servers. We do not maintain long-term backups of raw user biometrics.
- Mathematical Vector Storage: While the source photo is destroyed, the extracted geometric patterns, mathematical vectors, and mapped astrological identifiers are retained anonymously. These numerical coordinates are completely decoupled from your identity and are utilized strictly to train our computer vision algorithms to detect planetary lines with greater precision over time.
5. Data Security & Storage Architecture
Your data is encrypted both in transit (via TLS 1.3) and at rest (using AES-256 standards) across our distributed server network. We enforce strict role-based access control (RBAC) internally; engineering and research access to the production data environment is restricted to authorized personnel operating under rigorous audit logging.
6. Data Deletion and Grace Period
You retain the absolute right to purge your existence from our systems.
You may initiate a total account deletion either through the mobile app or our web Trust Center. Because data destruction is irreversible, we enforce a mandatory 7-Day Grace Period for security against malicious or accidental deletion. During this window, your account goes dormant but can be recovered. After exactly 7 days, your profile, authentication tokens, birth coordinates, and chat histories are permanently obliterated from our active servers.
7. Policy Modifications
As our AI architecture evolves, this Autonomous AI & Privacy Policy will be rigorously updated. Substantial shifts in data usage or training methodologies will be communicated via in-app notifications prior to implementation.
8. Corporate Contact
For all inquiries regarding data sovereignty, model training policies, or GDPR/CCPA compliance, direct your communication to our core office:
Email: support@vedicai.co.in