AI Case Study: Chat GPT Integration with OpenAI Models

6 May, 2023 |

How can you make a Chat GPT integration with OpenAI models into a software development successfully?


Technology advances by leaps and bounds and provides us with more solutions and possibilities to explore in the world of development, which can take us to unimagined places. The need to be constantly at the forefront of this range of possibilities, leads us to be in training and learning 24/7, which allows us to incorporate new expertise to, for example, integrate OpenAI models within projects with cutting-edge technology, such as a Chat GPT integration.

In this blog post we want to share with you how together with one of our large clients we have managed to implement a concrete business case where we made a Chat GPT integration into a custom software solution.

The objective of the application is to provide a dynamic and flexible training platform for the sales force of a renowned pharmaceutical laboratory, with the ability to obtain online information without the need to perform previous data uploads, saving costs and time.

The software solution, beyond including standard user, group and profile administration functionalities, contains modules related to training management: roles, suggested exams per role, exam form and results tracking per exam, per role and per group.

The important innovation we achieved is the integration with Chat GPT combining two of its main functionalities: Information Search and Text Analysis.

After a series of concept tests carried out by our team of Prompt Engineers together with business specialists on the client’s side to refine the parameters that allow us to obtain information in an accurate, reliable and fair way in terms of the amount of bytes sent and received to optimize costs, we concluded the following:

  • We use “Information Search” to obtain online information related to drug types, typical information contained in a drug package insert.
  • We use “Text Analysis” to compare the text of the information obtained versus the text of the answer entered by the user and according to the % of accuracy obtained we give a score to his answer.

The sum of your scores will give you a final result that is recorded and will be part of your training record through integration with your LMS (Learn Management System).

The results are amazing with a tremendous positive impact for the client in terms of cost and time due to the high degree of automation of the process for training your sales force.