2 min read

The OneBot tool from a technical perspective

Care to share?

The creation of the BERT natural language processing (NLP) model at the end of 2018 brought significant changes in eCommerce customer service. The ease of using BERT made it possible to make NLP technology more accessible and popular. That’s especially noticeable in the automation of call center systems and also in the increasing number of chatbot applications. 

The wide distribution of machine learning algorithms has allowed for the creation of tools that make the personalization of the customer experience more robust. Consequently, it’s helped facilitate the purchasing decisions of customers.

The technologies behind Onebot

The chatbot systems and personalization solutions are mainly in the B2C domain. However, with OneBot, we went beyond this framework in order to use new opportunities to improve the experience of B2B eCommerce customers. For these customers, the platform is also a workplace.
Configurators help to choose the appropriate product variant and can be defined with a minimum of development work. With the appearance of a new product assortment, it allows for the rapid creation of complex product builders as well as simpler variant filtering.

An individual approach

The personalization engine allows you to effectively propose new products when it actually makes sense for the customer. With OneBot, we chose to use the Personify XP.

Personify XP is the world’s first anonymous, real-time analytics and personalization platform. It automates digital experiences for online shopping through engaging, automated website personalization. It was named in Gartner’s 2019 “Cool Vendors in Personalization” report and drives an average 10% increase in revenue. 

The company helps their clients through their proprietary artificial intelligence (AI) by dynamically inserting content and products based on shopping context without the retailer or brand’s involvement. Identifying customer behaviors anonymously, the platform assists in creating new consumer engagement to create highly personalized experiences that convert. They partner with global brands that include boohoo, Farfetch, 7 For All Mankind, Pentland Brands, Hawes & Curtis, and Lovehoney.

Understanding language

The use of NLP in our solution allows clients to control accounts and orders. It allows the user to perform these tasks in the manner and order that suits the client without searching for the appropriate sections on the page. Customers can ask from a chat about the shipment, order history, or any other questions they might have.

Product expert

The independent connection with the product information management (PIM) system, NLP model, and more specifically, the Question-Answering function allows clients to ask questions about the product or confirm its features and attributes. The Q-A functionality is trained on the product data and can find sentences in its description that answers questions that are asked.

All in one place

The Team Activities section within the chat allows communication and access to activities across the entire purchasing team. The Activities stream helps to better synchronize tasks like creating an offer, delivery time selection, or even finding the best deals with active promotions.


The use of various technologies and tools allowed us to create a guided shopping platform for the B2B industry. Business customers also want to make purchases without the need to talk to a consultant, and with OneBot, they can get help at every step of the purchasing path. Here, you can learn more about the features of the tool and sign up for a demo. If you want to talk about the technologies used, just contact us.

Published October 3, 2021