Amazon, Microsoft and Google have all invested heavily in chatbot development tools on their cloud platforms. These...
tools make it easier for developers to create conversational interfaces for both consumer- and enterprise-oriented apps. But before you use them, it's important to understand the specific goals you want to achieve, as well as the different options available on the public cloud.
The chatbot development tools on Amazon Web Services (AWS), Microsoft Azure and Google Cloud Platform all support voice and text inquiries against a collection of back-end services.
Voice interfaces can include dedicated hardware, such as Amazon Alexa or Google Home devices, as well as smartphones or even Windows PCs that feature the Microsoft Cortana app. A company can also opt to use text-based interfaces and integrate them into websites to enhance application functionality.
"The general advantage of systems using natural language processing via text or voice -- and the reason so many deep-pocketed companies are investing in it -- is that it can be viewed as the next level of [an] operating system," said William Meisel, president of TMA Associates, a provider of consulting services for conversational interfaces. "This overcomes the limitations of the graphical user interface that has served us so well for so long, as it becomes overburdened with too many features and too many apps, particularly on mobile phones where the screen is smaller."
The use of chatbot development platforms is still in the early stages, as enterprises struggle to identify the best ways these tools can complement their web and mobile apps. For example, voice interfaces, in many cases, might work better for consumers at home than for workers in a busy office. Text-based interfaces, on the other hand, can improve customer support or even automate enterprise processes via integration with apps like Slack.
Compare chatbot options in the public cloud
Each of the chatbot development platforms available on the major public clouds has its own set of benefits.
William MeiselPresident, TMA Associates
Amazon Lex, for example, provides easy integration with AWS tools and resources and benefits from a growing install base of Alexa devices, as well as a significant library of Alexa Skills. Microsoft Bot Framework builds on the company's army of developers and also features strong integration with Cortana and Microsoft Cognitive Services. Meanwhile, Google's Dialogflow is based on technology from API.ai -- a company Google acquired in 2016 -- and, as a result, might provide the best cross-platform support of the bunch.
Here's a closer look at each:
AWS Lex, a service that includes capabilities for natural language understanding, builds on Amazon's core Alexa technology. Developers create Skills that essentially act as app components and can weave these Skills together to build more sophisticated chatbot interfaces.
Lex supports a variety of use cases, including virtual call center agents, information retrieval and enterprise productivity apps. Consumer-oriented Lex applications can run on over 10 million Alexa devices. For example, 1-800-Flowers has a Lex-based e-commerce app that gives consumers an alternative channel to buy things. But Amazon has also positioned Lex for enterprise use, such as in a call center to supplement live agents.
Developers can host their Lex applications on the AWS platform on a pay-per-use basis. Lex is priced at $0.04 per voice request and $0.00075 per text request.
Microsoft Bot Framework
Microsoft Bot Framework builds on Microsoft Azure, the company's extensive Cognitive Services library and the Cortana app. Developers can deploy chatbots on private servers or within Azure Bot Service. For greater scalability, bots can also run on Azure Functions.
Bots can interact with users via a variety of channels, including text/SMS, Slack, Facebook Messenger, Skype, Teams, Kik, Office 365 mail and other popular services. Developers can build them in C# or Node.js. Microsoft makes it easy to publicize new bots in its bot library, which also includes tools to call third-party bot services for specific functionality, like ordering a cab or querying an enterprise app.
Azure Bot Service includes a free tier, as well as a more advanced tier, called S1, that allows a chatbot to send unlimited messages to what Microsoft calls "premium channels" -- or a customer's custom-built apps -- at a rate of $0.50 per 1,000 messages.
Google's enterprise chatbot strategy revolves around Dialogflow, a rebranding of the API.ai platform it acquired. Developers can build apps using a variety of languages, including C++, Python, .NET, Ruby and Java. Development kits are also available for iOS, Android, Cordova, Xamarin and HTML.
Lauren Kunze, CEO of Pandorabots, says too much data, too little data and legacy systems can all be barriers to chatbot development.
Google continues to integrate the core Dialogflow functionality into its Google Assistant service for hardware and Android devices, its Chatbase platform for analytics and its various messaging services. This builds on Dialogflow's support for the widest variety of messaging formats and platforms, including Alexa, among the major cloud chatbot tools.
The Dialogflow platform currently includes 40 agents, or basic templates, that help developers build different types of chatbots. This makes it easier to customize an existing chatbot for a particular use case rather than develop one from scratch.
Dialogflow is available in a Standard Edition, which is free, and in an Enterprise Edition, which comes with additional features and costs $0.002 per text request and $0.0065 per voice request.