Kandy is a platform that makes it easy to embed communication features like video, chat, voice, presence (i.e., whether a user is online or not), screen sharing, and conferencing into any website.
Over the last decade or so, the world has moved on from the austere, forms-based approach of early web sites to sophisticated web applications. These applications use highly functional user-interface widgetry to provide an interactive responsive experience and seamlessly embed images, video, and audio into the user experience. However, some things are still quite hard to do. One of these is the area of functionality addressed by GENBAND Kandy: the integration of live, person-to-person communication within a website.
To understand the need for this product, consider the current premium support offering from a well-known customer relationship management (CRM) Software as a Service (SaaS) vendor. The SaaS application itself is fairly sophisticated, and within it are embedded support forums and training videos. However, the moment you want to communicate with a human being (an option that requires payment of additional maintenance fees), things get a bit less integrated. First, there is a page that, with premium support, allows you to raise a textual query. After an initial delay, it allows you to exchange messages with the support agent. At this point, you are still inside the SaaS application. However, once the support agent has done an initial evaluation of your ticket, the agent will typically ask you to share your screen by sending you the URL to an online meeting, which is set up on an external service from a well-known supplier. This needs to start up outside the application in a separate window, and it may even need to install itself first. Then, to make things even less well integrated, the agent doesn’t use the online audio on the meeting service—if the agent needs to speak to you, you get a message requesting your phone number, and you get a call.
It’s ugly. What you really need is just a little widget that sits inside the web application and links the communication between the application users and the support agent, sharing screen and audio as required. It should be as easy as embedding a Google Calendar widget. In fact, there is a range of different linked services that should be embedded this way, including video, chat, voice, presence (i.e., whether a user is online or not), screen sharing, and conferencing. At the moment, these tend to be accessible only via telecommunications infrastructure, which is either complex to set up and install or, in the case of online services like Skype, standalone and not easily embeddable.
This kind of requirement is fairly widespread in all sorts of different websites. It can apply in eCommerce applications where customers want product advice and in enterprise applications where users want support, and it can even become a core element of eLearning sites.
It is in this context that we were introduced to GENBAND Kandy. GENBAND is currently a global leader in smart networking solutions, SIP trunking, VoIP, unified communications, WebRTC, enterprise SBC, IMS, IPX, MGW… Clearly none of those telco acronyms make that much sense in the context of what we have just been discussing. Most organizations wouldn’t feel the need to set that stuff up in order to provide some additional features on their websites. GENBAND, however, is thinking of a future when people cease to consider voice communication as a standalone activity (analogous to picking up a phone); rather, they will expect to be able to communicate inside a web application seamlessly, as part of the application itself.
GENBAND has identified a route allowing it to package something equivalent to its core communications offering into an easily consumable form, which it has launched as “GENBAND Kandy.” This is a set of widgetry that embeds into web clients, authenticates with your application via OAuth, and can bind authenticated users together into communicating groups though SaaS delivered from the cloud.
GENBAND describes Kandy as a Platform as a Service (PaaS), but it doesn’t really fit into the classic PaaS model that readers of The Virtualization Practice may expect, e.g., Force.com, Heroku, OpenShift, or Cloud Foundry. Its feature set is almost entirely disjoint from those PaaS, in that it doesn’t allow general application deployment, and it addresses a set of problems that none of those PaaS can really help you with. The reason why GENBAND considers it a platform rather than an application service (like authentication) is that the various communications services are provided in a coordinated way.
There is no capital cost to running the Kandy infrastructure; Kandy is charged at a very small number of dollars per user per month. Since all the communication goes via Kandy, and none of it goes into and out of the enterprise’s own servers or cloud, there is no complex infrastructure for the enterprise to install and maintain. There is only the requirement to bind the application logic to the way Kandy is configured: for example, to set up “hunt groups” to ensure that support calls are routed to the next available support agent.