Boomi Makes Chatbots Better

7 minute read | 23 Jun 2019
Categories:

By Chris Cappetta

You no longer need to be a developer to create chatbots in-house. Using Boomi’s unified platform you can quickly and easily build and deploy bots to help you automate time-consuming processes and improve your customers’ experiences.

Gartner has posited that by 2021, over 50 percent of enterprises will put more of their yearly budget into bots and chatbot creation than they will into traditional mobile app development, thanks to their promising potential.

A chatbot is a program that simulates human conversational responses, and is primarily used to answer questions or perform automated services based upon questions and responses, helping streamline and improve customer support.

Chatbots are typically used via virtual assistants such as Amazon Alexa, via messaging apps such as Facebook Messenger or WeChat, through SMS or text messages, social media platforms like Twitter, or via individual organizations’ websites and applications. Using natural language processing (NLP), chatbots can recognize and understand human speech, infer context, and either retrieve information or take some other action, such as forwarding a more advanced request to a human being.

Types of Chatbot Platforms

There are two general types of chatbot platforms. Chatbots are often sold as a bundled vertical-specific package. They work well for the use case they were designed for, and they are quick to stand up. But they are not necessarily versatile — you can’t take a chatbot that was designed for the auto industry and use it in manufacturing, for example.

The other option is to use an artificial intelligence (AI) chatbot development platform that offers NLP capabilities, like DialogflowAmazon Lex or Salesforce Einstein. AI is very strong at recognizing the intent of the question, handling synonyms and handling different ways to phrase the question. But if you need to craft responses using data from multiple systems — perhaps based on keywords — it becomes a coding exercise.

Boomi Builds Bots

Last year, a customer came to Boomi with a request for an AI chatbot embedded in its website that would help it identify and route leads. We had already built some rule-based chatbots, so this was a great opportunity to take our technology to the next level. Using Boomi Integration and Boomi Flow, the team was able to quickly put together a proof of concept based on a pattern that combines AI with business rules.

The customer was thrilled with our chatbot solution, telling us that it received a full return on investment in the bot from the very first lead converted.

Now, its marketing team knows which leads are hot: The AI identifies people who are likely to be in the decision-making stage based on their engagements with the website. If the business rules recognize there is a short timeline for purchasing, it can send a message to those identified leads that says, “Hi, it seems like you’re ready to make a purchase. Can we have one of our sales reps reach out to you?”

Read our blog post “Low-Code Integration Helps Hardworking Bots Earn Their Keep” for more on how Boomi makes bots better.

An Integration Layer for Speed and Versatility

An integration layer like Boomi’s that can connect machines and humans saves you development effort and time. Boomi gives you the speed of a vertical-specific chatbot with the versatility of a development platform. As Boomi primarily uses configuration rather than code to achieve the integrations, you don’t need a developer to create simple chatbots.

You can use the AI platform of your choice, but at the point that you would begin coding to support dynamic integration or specific rule-based responses, you instead pass the keywords and data to Boomi. Boomi takes care of configuring those integrations and providing the data based on logic and rules for the kind of info the chatbot needs to supply. From there, the Boomi Platform can help the chatbot craft the responses to the user.

Boomi identifies the correct application or endpoint the bot needs to access. Our rules-based integration pattern uses artificial intelligence to craft a response for the user, effectively engaging the human layer. It uses AI for NLP; recognizing what a person is asking, handling synonyms, etc., so it can understand the bot’s request, connect to the right application, fetch the information, render it in the proper format and send the right response to the requester in a matter of seconds.

Pervasive Connectivity and Quality

Boomi also gives you versatility by offering multiple services that can be used to meet your specific chatbot needs. Each of the services available within Boomi’s unified platform strengthens the others, giving our customers the ability to do more, faster, more efficiently and more effectively.

If you need a bot that is engaged by humans through existing systems and technology rather than a new embedded web offering, like a text message or instant message, you can build that using Boomi Integration. If you need a human interface, you can build your bot and present the responses in Boomi Flow.

Boomi Master Data Hub will help you ensure the data you are accessing and presenting back for responses is correct and up-to-date. And you can extend the bot’s reach using Boomi API Management.

For example, someone could type a message in Slack that could trigger a number of different integrations using APIs which send the response back to the channel of choice, whether that be Slack, SMS, voice or something else. That’s configured into that particular pattern.

This is just the beginning of how Boomi can help you build better chatbots – or any bots, for that matter. We are currently developing new frameworks and other tools to help your organization build bots better so they can do more. Stay tuned!

To find out more about how Boomi can help you build chatbots and take advantage of other emerging technologies, contact our Boomi integration experts today.