Loading...
Please wait a moment
Loading...
Please wait a moment
California AB 2013 Compliance
Last updated: May 10, 2026
Coverforme LLC provides generative AI services for voice transcription and summarization. We are committed to extreme transparency regarding how your data is processed.
Our services are powered by models and technology from third-party providers across several categories: large language model providers, voice infrastructure providers, speech-to-text providers, and web content extraction providers. The current named list of these sub-processors and their roles is maintained on our Subprocessors page and is updated whenever we add or change a sub-processor.
Coverforme LLC does not "train" or "fine-tune" foundational artificial intelligence models. Our technology is built on a "Zero-Data Retention" (ZDR) architecture for all third-party AI processing. We do not use call transcripts, audio streams, or metadata to modify the weights, biases, or performance of the underlying AI models provided by our sub-processors.
Our "modifications" are limited to Proprietary Prompt Engineering. This means we provide the AI with specific, real-time instructions and structural templates to ensure your summaries are accurate and professional. These instructions are "stateless," meaning the AI uses them to process your specific request and then immediately "forgets" the data.
Coverforme LLC does not utilize synthetic data generation in the development or modification of our service interfaces.
We exclusively use the enterprise-grade or business API tiers of our model and voice infrastructure providers. Unlike consumer versions of these tools (such as free-of-charge consumer chatbots), our API-level agreements with each provider ensure that your data is not used for model training by default.
No Audio Recording: Calls handled by Coverforme are processed by AI in real time. We do not record or store raw audio. A written transcript and AI-generated summary are saved in the User's private dashboard so they can review what callers asked for. The underlying audio stream is discarded immediately after transcription and is never stored in an audio format or used to train any AI model.
When the Services are used for food-service, hospitality, or other verticals that rely on uploaded or configured reference materials, AI-generated answers may still be wrong, incomplete, or outdated — including responses about menu items, ingredients, allergens, pricing, promotions, and whether a dish or offer is available — even when those answers are grounded in files or facts the User provided. Callers should treat allergen and severe dietary questions as requiring confirmation from on-site staff when safety matters; the assistant does not provide medical, nutritional, or food-safety advice. For the general accuracy limitation, see AI Processing and Accuracy Disclaimer in our Privacy Policy.
For more information on the logic and intended output of our AI systems, please refer to the Automated Decision-Making Technology (ADMT) section of our Privacy Policy.
For documentation on the underlying models and their training data, see each provider's public model documentation. The current named list of model providers we use is published on our Subprocessors page, where each provider links to its own model documentation.