Capacity's Approach to AI

Capacity utilizes various AI tools and models to power its platform. For years, we've used a mixture of open source and in-house AI models to power our Natural Language Processing (NLP) pipeline that allows our chatbot to understand the intent of incoming requests and messages in order to provide more relevant answers. We've also used AI models to perform Intelligent Document Processing (IDP) and extract useful information from documents and forms submitted to us from customers.  Capacity has reviewed these tools and models to confirm they perform as expected and all of them are hosted by us (in other words, Customer Data is not processed outside of the Capacity environment in utilizing these AI tools and models).

More recently, we've started using Large Language Models (LLMs) in certain aspects of the Capacity platform to provide a better understanding of language and the ability to generate natural-sounding sentences.  A generative pre-trained transformer (GPT) then can take these sentences and customize them according to our subscribers' needs. 

Capacity currently uses OpenAI’s large language model (LLM) model for our generative pre-trained transformer (GPT).  OpenAI API is SOC 2 Type 2 compliant.  (Note that OpenAI API is different than OpenAI's Non-API consumer services like ChatGPT.) 

    • By default, pursuant to Open AI policy, OpenAI will not use data submitted by customers via its API to train OpenAI models or improve OpenAI’s service offering.

    • OpenAI’s policy is that any data processed from Capacity to Open AI will be retained for abuse and misuse monitoring purposes by Open AI for a maximum of 30 days, after which it will be deleted (unless otherwise required by law).

  • For more information on OpenAI’s policies, please see https://openai.com/policies/api-data-usage-policies

By default, Capacity's Helpdesk AI Tools which utilize OpenAI is disabled in Capacity’s platform for subscribers (a/k/a our customers) and can only be toggled on by an administrator of our subscriber.

Capacity recommends that subscribers prohibit the use of this feature where protected health information is included in Subscriber Data or where our subscriber is covered entity or business associate under HIPAA.

All data that is processed using Capacity's AI Toolkit is stored in the US.

Was this article helpful?