Omnichannel for Customer Service and Power Virtual Agents

I have recently been experimenting with both Omnichannel for Customer Service and Power Virtual Agents, as now the two can be a combined their integration seemed like an obvious thing for me to review. In this post I will document how I created my first “omnichannel virtual agent” and make a few observations along the way.

A good starting point might be for me to describe the use case for Power Virtual Agents that can hand off to a human agent. BOTs are great at answering commonly occurring questions as it is pretty easy to train them with suitable topics and responses. However real agents will always remain the best people to answer unusual questions or give responses when a judgement call is needed. In fact I personally consider a human hand off option as an almost essential feature! Without it your BOTs offer an intelligent conversation capability but when they “run out of steam” your potential customers are left in a frustrating dead end. Implementing an escalation from the BOT to a human avoids customers reaching this frustrating state.

Integrating Power Virtual Agents with Omnichannel for Customer Service allows us to define multiple queues and transfer customers from our AI friends to humans in an almost seamless manner.

In the video below I have given a demonstration of how Power Virtual Agents operate, how to create the virtual agent and how to configure Omnichannel for Customer Service to use your BOT.

Conclusion / Limitations

Overall I’m very happy and excited about the integration of Omnichannel for Customer Service and Power Virtual Agents. I’m already thinking about using this with one of my customers!

I am very happy with the virtual agent handoff, so I don’t want to be overly critical but I have seen a couple of limitations (with possible workarounds);

One “limitation” I’ve observed is that the Power Virtual Agent’s transfer option doesn’t link to opening hours …. I’ve often seen a BOT enabled out of hours, designed to help customers whilst the agents aren’t working. During these “quiet” hours we wouldn’t want the BOT to attempt to transfer to the absent agents. Out of the box I haven’t seen a no code approach to this scenario. I guess it would be possible to create two virtual agents, one that will transfer to a human and another that doesn’t. You’d then need a small piece of code in your website to decide which BOT should be active at any point. This approach would be pretty simple to implement but I’m looking for my implementation to be no code, hence why I see this as a limitation!

BOTs do not support skills based routing. But what if we want to apply skills based routing when transferring to our human agents? In my video demo I showed that we can use the “va_Scope” context variable to allow us to know when to route to a queue of humans. I guess we can simply ensure that the humans in that queue have the skills to handle any query from that Virtual Agent. But I would have liked to have seen skills based routing, maybe with the transfer request being associated with multiple Omnichannel for Customer Service skills. We might not have skills based routing from BOTs to humans but all is not lost! As the tile used in my virtual agent does include the ability to define a private message to the human agent, that might prove really useful! It wouldn’t take much effort to route particular chats to various queues based on the private message from the BOT. We’d use the context variable “va_AgentMessage”, which I believe would include the private message from the BOT.

Knowing what the possible context variables are might be useful! So below you can see a summary of all of the context variables available.

I hope this post and video will have been useful to anyone considering using Power Virtual Agents integrated with Omnichannel for Customer Service …. Enjoy!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s