Security

This page outlines our security posture and practices. ThinklytixAI is an AI-powered analytics and conversational intelligence platform built to protect data privacy, security, and client control.

We never use client data to train, fine-tune, or improve any Large Language Model (LLM) — public or private. All AI interactions occur in real-time using secure, ephemeral processing; data is never persisted outside the client’s environment.

Standards & Compliance

• SOC 2 readiness and best practices.
• Encryption in transit (TLS) and at rest where applicable.
• Access controls based on least privilege and role-based authorization.

Protecting Your Data

• Network segmentation and monitoring.
• Vulnerability scanning and timely patching.
• Routine backups and recovery testing.

Core Principles

1. Data Ownership – Clients fully own their raw data, derived datasets, and generated insights.
2. No Model Training on Client Data – LLMs operate only in stateless inference mode; no retention, fine-tuning, or reuse occurs.
3. No Cross-Client Data Exposure – Queries and responses are isolated per client.
4. Controlled Access – Access governed by RBAC and MFA.
5. Transparency – Clients may review processing logs and configurations.

Responsible Disclosure

If you believe you’ve found a security issue, please contact us at security@thinklytix.ai. We appreciate responsible disclosure.