How is Creatio integrating ChatGPT into its Atlas low-code platform?

The integration of the ChatGPT connector into Creatio’s offerings will allow developers to build applications faster and help in customer-centric use cases such as providing assistance in writing emails and supporting customer support executives.


Workflow automation provider Creatio is integrating a ChatGPT connector into its Atlas low-code platform and customer relationship management (CRM) products, the company said on Thursday.

“In order to help enterprises achieve more efficiency and productivity, Creatio is integrating ChatGPT with its line of products. The integration will be included in our Atlas platform and it will provide additional support for other customer-centric use cases via other offerings,” said Andie Dovgan, chief growth officer at Creatio.

The ChatGPT connector will be embedded within the Atlas platform interface as a prompt and developers will be able to access it for generating templates for building an application, Dovgan added.

The connector, according to Dovgan, will help enterprises develop applications faster as it prepopulates the data model when asked to generate a template for building applications.

“Once the data model is showcased in the low-code platform, developers can follow a drag-and-drop journey to complete the application and publish it,” Dovgan said.

How does the ChatGPT connector work?

After the developer or the user has engaged the prompt of the low-code platform, it asks the user to define the purpose of the application and the type of data (i.e. customer data, sales data, etc.) it will work within natural language, a spokesperson from Creatio’s R&D unit said.

Before the integration of ChatGPT, developers would have to follow a wizard to do the same task.

Once the natural language query is received by the connector, it structures the prompt in a way that OpenAI can understand and provide a consistent JSON response format. A prompt is then sent to the OpenAI API endpoint, which in turn receives an answer from OpenAI in JSON format.

Once the JSON response, which carries a description of the data objects, fields, data types, and data relationships, is received by the connector, it verifies the response and ensures that the expected format and data sought in the query is followed, according to the spokesperson.

“The connector then transforms the JSON response to an application using a predefined set of rules and applies the required template layout to the application by asking the developer before publishing the application,” the spokesperson said.

Creatio to integrate ChatGPT connector into emails, chat

Other than the Atlas platform, Creatio is planning to integrate the ChatGPT connector into its other offerings, such as CRM products, in order to support customer use cases.

The connector will be integrated into the user interface of various product offerings, the company said, adding that the GPT prompt will be integrated into what it calls “communication panels”, which include customer-facing or communication interfaces such as emails, chat, and phones.

“To give an example, if I'm a salesperson, and I have emails that I want to send. I can ask ChatGPT to put together a template for me and then I can copy-paste that template into my email and send it. The other example could be ChatGPT guiding a customer support executive while the person is talking or chatting with a customer,” Dovgan said, adding that the GPT engine can provide suggestions based on the conversation.

Copyright © 2023 IDG Communications, Inc.