LiveChat has announced its integration with BotEngine, a chatbot platform. From now, each of over 21,000 companies already using the chat application is able to boost their customer service with bots with just one-click.
Customers gladly turn on the LiveChat widget when looking for customer service on digital platforms. This is proved by the fact that creators of LiveChat are monitoring the activity of over 1 billion end-users visiting its customers’ websites every month. What they mostly appreciate in customer service via chat application is its speed and effectiveness.
Customer service is moving quickly with evolving technology. Currently, bots are a hot trend. Some people might not want to chat with robots at first, but it changes when bots give them what they expect – convenience, speed, and personalization.
Mariusz Cieply, CEO at LiveChat
LiveChat offers an integration with bot-building service to help companies in making their customer service accessible and productive. BotEngine is an advanced tool which allows designing a complex AI-powered communication strategy.
All bot scenarios are created directly in BotEngine as stories, which later are triggered every time a visitor starts a chat on a website. Bots’ chats are tagged and stored in the LiveChat archives. In this way, it’s easy to filter them and set apart from the ones conducted by agents.
Moreover, BotEngine archive allows a user to take more advanced information about bot’s conversation. In the platform, a user has access to text analysis and can easily detect if the bot’s answers are appropriate or not. All incongruities are listed with an accuracy confidence score helping the user to improve the bot scenario. Conversations with bots can also be rated by visitors with thumbs up or down at any moment. Monitoring customer satisfaction helps to better understand the bot performance.
The platform uses Natural Language Understanding. This mechanism receives data from a customer’s query and tries to find the most suitable answer based on collected information and current user path in the bot scenario. When the user is on the path between two similar interactions, then the algorithm first tries to match user input message with the next step of the story. While creating a bot scenario, a user can add ‘entities’ and define a confidence score that will be used in their matching. In this way, even when a visitor makes some mistakes, e.g. instead of ‘New York’ writes ‘New Yorr’, a bot will be able to match correct answer put in a scenario.
If a bot can’t complete a task, the conversation will be forwarded to an agent. It might happen in two cases:
- when visitor exhausts a defined limit of questions which are not included in the scenario;
- when a user creates a scenario with a specific ‘action’ which automatically triggers forwarding to an agent.
LiveChat rolled out BotEngine as an independent product in July this year. Since the launch users created over 1,000 bots’ scenarios. From all feedback, the creators have received, what users appreciate the most is platform’s UX and UI.
Users feedback has a huge value for us as our further aim is to build a marketplace around BotEngine where bot makers will be able to share their knowledge, experience and most importantly – the results of their work: chatbots. We want to develop the platform in ways that are mostly expected by its users.
Dariusz Zabrzenski, Head of Research and Development at LiveChat
All instructions related to the integration are described in tutorials. In the next few weeks, the ready-to-use bots will be able to be implemented directly from the platform’s marketplace. The bots scenarios will be mainly based on lead generations, sales or in-house usage for different types of business. The creators will also encourage bot makers to share their own scenarios in the BotEngine marketplace.
Boston, Massachusetts — 9 October 2017
