Today’s customers have high expectations from brands. They expect fast and personalized service across all digital channels — service that expertly anticipates a customer’s needs and solves their problems. This expectation also applies to their interactions with chatbots.
Unfortunately, most of today’s chatbots aren’t up to the task. And it’s all because of how they’re built.
Advances in natural language processing have made huge strides in recent years — promising a future where chatbots are easier to build, offer more human-like conversations, and produce better customer interactions.
Unfortunately, the bulk of chatbots built today still use only intent-based approaches within their NLU engine. Intents don’t do a great job in answering questions, as they miss context, nuances, and meaning from the user utterances. Plus, intent-based engines don’t scale — the addition of more intents directly impacts the performance of the entire model. As a result, most chatbots fall short when dealing with unexpected questions, leading to a customer experience that is far from outstanding.
It’s time to change this.
It’s time to say goodbye to intents, and build better chatbots, that handle more questions and topics, and that are faster and easier to build. This is why we created OpenBook by Botpress, the world’s first commercial knowledge-based NLU engine that actually works.
OpenBook brings game-changing benefits to the world of chatbot development.
Build faster & easier
OpenBook helps developers build faster and easier by entirely scrapping the intent-based approach. No longer hindered by intents, utterances, and confusion matrices, developers simply have to add facts, deploy, and allow OpenBook to supply the answers to all the user's questions.
Predictability & control
Predictability is essential when building a chatbot — conversations can easily go rogue and leave the developer with little control. Thanks to a selection of helpful settings, OpenBook gives developers full control over their conversation flows, enabling complete predictability and easy management of preferences.
Delivering conversations at scale — that’s one of the biggest goals of any chatbot development.With intent-based bots, performance can suffer with each new intent, and when adding new information post-production. With OpenBook, however, developers can add new facts and information without harming performance.
Human conversations = better CX
Chatbots are often criticized for delivering robotic and canned responses — resulting in clunky conversations that diminish CX. OpenBook offers a revolutionary breakthrough in quality, creating conversations that feel authentically human.
We asked several expert chatbot builders to help us answer this question. We tasked each to build comparable chatbots with Rasa, Google Dialogflow, IBM Watson, and OpenBook. We then asked users to rate their chat experiences using each test chatbot. We got remarkable results! OpenBook powers conversations that are more than 2X as accurate, builders were able to create chatbots 40X faster, and OpenBook chatbots are able to handle 5X as many questions. Read the details of our study — and get access to the data set–in our OpenBook whitepaper.
OpenBook is still in development! You can try the OpenBook demo here and get a feel for how powerful it is.
Interested in learning more? Be the first to learn more about OpenBook, and sign up for access to the private beta when it opens in a few weeks.
Learn more about the capabilities of ChatGPT, including its capacity to understand multiple languages.
Find out how to add a chatbot on your website as well as how to test it.
Dive into our new video series where we build a ChatGPT recipe chatbot from scratch.