As an AI transformation partner, Lost Relic adheres to the strictest privacy standards when designing and orchestrating autonomous agents for our clients.
Zero Public LLM Training
The most critical pillar of our privacy framework is that data processed by the AI agents we build for you is never used to train public Large Language Models (LLMs). We utilize enterprise-grade API endpoints with foundational model providers (such as Anthropic and OpenAI) that contain explicit, legally binding guarantees against utilizing telemetry or prompt data for model training.
Access Control
We operate on a principle of principle of least privilege. AI agents are granted access exclusively to the specific tools, directories, and data points required to execute their specific automated tasks, reducing the attack surface and preventing unintended data exposure.