Enterprise AI in One Business Day: A case study 

The Client, a Fortune 500 corporation, wanted to utilize OpenAI’s ChatGPT service so its employees and customers could access, analyze, and compile proprietary information using the service’s renowned system of user-friendly, conversational prompts.

However, because it was a large, publicly traded company, the security of its proprietary data was paramount; any implementation would need to protect its data, prompts, and responses. Fortunately, Microsoft’s new Azure OpenAI Service could allow them to create an isolated environment of OpenAI and maintain enterprise security.

Using this Enterprise AI service, the Client could place content into selected folders on its own cloud servers, allow the Azure OpenAI Service to index them, and then send prompts and responses to users without exposing the data to anyone outside the corporation.

The Client had an extensive Microsoft Azure relationship, but its in-house engineers had no experience with the OpenAI system. Eager to move forward quickly with an implementation so they could begin designing both internal- and customer-facing products, they asked Tricon Infotech to assist.

Tricon Infotech’s engineers, who were already familiar with the system, quickly navigated the Azure system to request a new OpenAI account for the Client. By Friday, they received the OpenAI credentials. Then they customized an existing Azure template (especially the front-end) and pushed a CI/CD deployment to the Azure Cloud.

By Monday, the application was live.

The Client now had a private, enterprise ChatGPT instance that was available only to its employees, customers, or other designated users, would utilize only documents that it approved, and keep all the prompts, responses, and data private to the Client. The new instance forms the basis for the development and release of new, innovative AI-powered applications.