Dynamics 365 Contact Center Explained with Alan Ross, Partner Group PM Manager

Alan Ross, Partner Group PM Manager, Dynamics 365 Contact Center at Microsoft delves into Dynamics 365 Contact Center and the role of Copilot in improving agent productivity and customer experience.

• Features and capabilities of Dynamics 365 Contact Center
• Decoupling from Dynamics 365 customer service
• How Copilot enhances the agent experience by providing context-aware suggestions, knowledge search and conversation summaries
• How Dynamics 365 Contact Center relates to/integrates with Microsoft Teams 
• Customer success stories and future developments in Dynamics 365 Contact Center

Thanks to Ribbon, this episode's sponsor, for your continued support and helping to make content like this possible.

Transcript

Copy Transcript Copied

Alan Ross: We've picked the best of the capabilities that align to a contact center and coupled it all together. Plus made it so that it is, uh, as we described fully composable or decomposable from the CRM.

Tom Arbuthnot: Welcome back to the Teams Insider Podcast. Got a special one this week, a real deep dive on Dynamics 365 Contact Center. Alan Ross has joined me, who is Group PM Manager for Dynamics 365 Contact Center. He leads some of the engineering teams there. And we did a really great deep dive on Dynamics 365 Contact Center, understanding what it is, what it isn't, how it works, what the licensing is, how it relates to Dynamics Contact Center Platform, Dynamics Customer Service.

Alan was great to explain how that world all hangs together and also where Microsoft are going with Dynamics 365 Contact Center. Lots of AI conversation as you would expect. Many thanks to Alan for taking the time to join the pod and explain it all. And many thanks to Ribbon, the sponsor of this podcast.

Really appreciate all their support of Empowering.Cloud on with the show. Hi everybody. Welcome back to the podcast. Been excited about this one. I've got Alan Ross on and we're going to go deep dive on the Dynamics 365 Contact Center. Really exciting area. I got a lot of call outs in the recent, uh, financial earnings call as well as a, as a big thing to look for in FY25.

Um, before I go further, Alan, do you just want to introduce yourself, please? 

Alan Ross: Sure. Thanks, Tom. Thanks for having me on the podcast. Uh, Alan Ross, I lead a team of product managers in the Dynamics Engineering Group focused on customer experience. We have a special focus on the new Dynamics 365 Contact Center SKU, um, but we look across the customer experience portfolio.

So, uh, we dabble in customer insights, D365 sales, and some other areas as well. 

Tom Arbuthnot: Awesome. Well, thanks for making time. Really appreciate it. As I was just saying in our, in our pre conversation, I'm not a, I'm more of the modern work team side of the house than the Dynamics side. So I'm going to ask all the, uh, all the questions to help everybody understand it.

Hopefully. Um, let's start off 

Alan Ross: with no problem at all. And I, I, uh, I started off, uh, my career at Microsoft focused on Office, uh, before Hopping over to Azure and then finally jumping over, uh, to business applications. So hopefully I can be a voice of a decoder ring where appropriate for somebody who's not necessarily super steeped in context.

Tom Arbuthnot: So let's start with, uh, what is. Dynamics 365 Contact Center, and then we'll work back to how it relates to the other products. 

Alan Ross: Sure. So, um, within the Dynamics 365 customer service portfolio, we've long had, um, a omni channel add on, which had both digital, social, voice channels. which was available as an add on to customer service enterprise.

Um, and you could use that for your contact center solution. In fact, we have many customers that we're doing just that. Um, what we've done since then, it was, we've of course, uh, doubled down on all things Copilot. Uh, we've brought in our nuance acquisition, uh, into our first party offerings, and we have ensured that the.

Offering is not tightly coupled with the Dynamics 365 Customer Service. Offering anymore. So, um, it works with third party CRMs in addition to the, uh, first party, uh, which of course you've, you've been able to get it. Uh, but now we've included all of that as a, as a single bundle offering. 

Tom Arbuthnot: Awesome. Yeah.

That decoupling to a huge deal, obviously, because that opens up the, the potential for customers to use it, who might be Salesforce houses or another house and they don't have to be a Dynamics house to take up a Dynamics 365 contact center. 

Alan Ross: Indeed. In fact, one of the biggest areas of demand that we're seeing is from customers that are on homegrown CRM.

So they've, uh, over the years built their own CRM on, you know, basic SQL plus middle tier, whatever that combination is. Uh, and, and they want to add channels to it and, uh, have a solution that works there. Um, that's one of the areas that we're, we're certainly supporting as well. 

Tom Arbuthnot: Awesome. And so how does this relate to, we had Dynamics Contact Center Platform before, how does this relate to that?

Alan Ross: Yeah, so, um, the way that we were representing, uh, DCCP, Uh, was really saying, Hey, we have a whole bunch of things within the Microsoft umbrella. You can pick and choose those and bring those together. Um, this is really a  dedicated SKU that we've picked the best of the capabilities that align to a contact center and coupled it all together, plus made it so that it is.

Uh, as we described fully composable or decomposable from the CRM. So as opposed to sort of a marketing term that included a whole bunch of different pieces of the portfolio, we've now said, Hey, here's a SKU and, uh, it's got a price and, uh, you can buy it and get, get what you need. Yeah. I was 

Tom Arbuthnot: excited to see SKUs.

We'll get onto SKUs and licensing in a minute, but like actual SKUs for an actual service. DCCP was really interesting because you had nuance in there. You had Power Platform in there. You had ACS in there, but it was a bit of a, uh, Uh, a set of building blocks that a partner could then bring together. You couldn't directly compare it to another kind of, when you're talking to a contact center team, I wasn't able to say, here's what it is.

Alan Ross: We, we heard, we heard the feedback loud and clear, and, uh, we, we, we believe we have addressed it with this new 

Tom Arbuthnot: offering. So this supersedes DCCP. So DCCP is no more and Dynamics 365 Contact Center is the play. 

Alan Ross: You will certainly, uh, start to see the DCCP marketing, uh, start to go away. Uh, in full favor of Contact Center.

Tom Arbuthnot: So you mentioned Dynamics, uh, customer service there as well. So help me understand what is that and how does that relate to Dynamics 365 Contact Center? 

Alan Ross: Sure. Um, so, so the way to think about the underlying CRM, the customer engagement CRM, that is Dynamics 365. Customer Service. You can get that, uh, in an enterprise version.

Uh, as I said previously, we, we had the enterprise version with an add on that you could get to get those channels. Um, we've now decoupled it. So the idea is, whether you're using our CRM, whether you're using someone else's CRM, whether you're using no CRM, so we even support headless. Uh, Contact Center, the Dynamics 365 Contact Center offering runs in parallel with those three options.

So, uh, whether you're doing it side by side or rotating it 90 degrees, uh, you can think about the Contact Center as kind of layering on top of any of the, those CRMs or frankly, any combination of those CRMs. So that's the other pattern that we see within larger enterprises. They say, Oh, we've got a sales.

Pocket of Salesforce over here. We're using some Dynamics over here. We've got these third party systems. We say, great. Dataverse is still the place where all of that comes together. We have the fantastic ecosystem of the 1, 400 different connectors, the third party systems. So we support all of that, um, as this overlay, which, Provides the channels, provides, uh, the agent desktop, whether that's, um, uh, digital agents, voice agents, blended agents, you get the, the desktop, you can get, uh, self service, so you can get the IVR, the chatbots, etc.

Uh, chatbots you do, uh, purchase through Copilot Studio, um, but we provide all of it, uh, in a single, in a single SKU. Microsoft Teams Uh, where there's only a couple other pieces you, you might want to add to get the complete context. 

Tom Arbuthnot: Awesome. So Dynamics 365 Customer Service is the CRM and I could have that and it would integrate, but I could also not have that.

I don't have to have that to buy Dynamics 365 Contact Center. 

Alan Ross: Correct. So we have, uh, customers that just want to be able to Manage cases, for example, and they're not ready to use our channels. Uh, we support other third party channel providers to embed the controls inside of our CRM or otherwise use their desktop with our CRM as the backend.

We still support all of that. Um, this is just the inverse of that, which is now we have our channels that are available both for our own CRM, as well as. For other third parties here. 

Tom Arbuthnot: Yeah, that's what I've seen in the past. I've seen Dynamics Customer Service and some other big contacts into kind of blending in, but it was the opposite.

That makes sense. And you, you touched on the different types of. Agents and capabilities there. Let's dig in on that. You mentioned the, I think, digital agent and voice agent. Is that the two? 

Alan Ross: Yeah, indeed. So, um, sort of the thing that seems to stand out on our, uh, agent desktop is a, the fact that we have, uh, unified experience.

So an inbound message, whether it's coming from social, digital voice, All shows up in a unified inbox. It's a unified experience. We enable transcription and translation of the voice calls. So from even a user interface for the agent, it kind of shows up in the same exact way, just a set of text representing that conversation.

The second piece is the Copilot capabilities, which are infused inside of that interface. So, um, Again, with the, with the consolidation of the channels into one place, it means that the Copilot experience is the same as well. So we have context aware Copilot. So our Copilot is sort of listening in on that conversation.

So it's able to suggest prompts, for example, it's able to connect to knowledge, whether that knowledge be in third parties or in our own knowledge source, you're able to get that full capability, the summarization of the chat, for example. Uh, or the voice call. So let's say, uh, the end user has had a conversation with a chatbot, or they've had a conversation with an IVR.

They've had a bunch of back and forths, and then at the point at which it escalates to the agent, now the agent is seeing the full transcript of the conversation that's happened up until that moment. Copilot can summarize the conversation that's happened. So if it's a long running conversation, they might want just a couple of sentences to say what's happened in that conversation.

Again, this is the kind of thing, having that be powered by generative AI and Copilot really seems to stand out, um, in the customers that we, have been, uh, using this set of capabilities with, uh, and then the sidecar experience for Copilot, where they can just kind of ask natural language questions. They can get it to refine the answers that come back from Copilot.

And it's really taking advantage of. Where the agent used to have to go look across a whole bunch of different sources for something. Now you expose those sources to Copilot, Copilot will look across them, come up with the most likely answer, uh, and provide that as a suggestion, which then, uh, creates type of impact on agents, uh, productivity that, that we've demonstrated.

That's 

Tom Arbuthnot: awesome. I really want to use that Copilot conversation. Just before I do, the agent experience, that's all web based. It's the, that's the experience for the, for the kind of agent client. Microsoft 

Alan Ross: Correct. Um, so it's still sort of, uh, built on the same, uh, customer service and workspace foundation.

Uh, we just now allow that to operate without the CRM underneath. 

Tom Arbuthnot: Awesome. So Copilot, AI, I'm, I'm super bullish about AI and contact center. Like I can't see how it can't have a huge impact. It's one of those areas where both it can. Save costs and improve experience. It's super rare in technology to have both those here.

Normally it's like, we can make this better, but it's going to cost you, or we can make this, you know, cheaper, but you're, you're, you're slightly impacting experience. I feel like AI and Gen AI, that the potential is to, you know, it's very hard to contact centers are tough to hire for, tough to train for, that there's so much opportunity there.

Um, I'm, I'm, I'm guessing with Microsoft, you know, pivoting so hard to AI, that's central to kind of your thinking there? Uh, 

Alan Ross: a hundred percent. And, uh, I mentioned this as sort of a differentiator for us, having the Copilot capabilities be baked into that agent experience. So there's not a separate SKU that you need to go buy in order to get Copilot for the agent, uh, comes right with it.

Uh, I already sort of, uh, teased some of the capabilities that come with it. But really, what we've done is we've looked across the spectrum of activities that the agent engages in, whether it's the getting caught up on a case, whether it's helping to close the case, whether it's getting information that helps to resolve the case, all of those places in which an agent has some friction in which they increase the average handle time, all the metrics that, you know, The leaders from contact centers have told us are important.

We've taken the opportunity with Copilot. Where is it? The generative AI can have an outsized impact on those metrics. That's where we've really focused on our investments in the innovations that we have really just looking at. The full scope of an agent experience in taking a case and moving it to 

Tom Arbuthnot: Clojure.

I've seen in some of the marketing, you can help me understand what's, what's here today. Some of the marketing was like, uh, backend agent assist type use cases. And some of the marketing was customer facing AI, doing the chat AI, answering the call. Are those. Both available today. Is the customer facing stuff kind of more vision where that's it?

Alan Ross: Yeah, sure. So, um, the bulk of the capabilities that we have out of the box are within the agent desktop experience. Uh, so combination of the summary, when you go close a case, we can take that summary and write it to the case notes. Uh, that's all generative AI powered. Um, the knowledge search capability is also one of the hero scenarios there.

So expose those knowledge sources to the sidecar Copilot. The agent can have a natural language conversation with it. It supports multi turn conversation, all of that baked in. Um, we do things like, uh, timeline summary. So if it's a long running case with multiple touch points, uh, from that particular uh, and customer, um, we can look across the entirety of the, of that history.

We can actually roll up a generative AI summary. And it's one of the, one, one of the most popular features and it's sort of no regrets, right? Cause we're not, Using that to then expose Gen AI directly to the end customer. So customers that are risk averse, they just want to say, where can I get quick wins?

That's one of the best places that we're able to show it. And, uh, we hear that feedback from, from agents as well. Um, in troubleshooting the service. My, my team and myself, even we use this capability. So, uh, it's great when we get pulled into some sort of escalation on something and we need to get quickly caught up to speed.

I've experienced the same thing where it's like, wow, this is a really long, long running case by the time it gets escalated to me. So great to get the top line summary right up front. So, um, that's all on the agent side. Uh, on the self service side, uh, as I mentioned on the chat bot, we have, uh, Copilot Studio.

Copilot Studio has, uh, GA'd so far, um, uh, Generative Answers. So that enables, just like you can expose knowledge to the agent, you can expose knowledge directly to the end user, uh, and, uh, uh, Uh, generative answers enables the lookup to those, uh, knowledge sources and a summary of the most likely response back, uh, that is available.

And it's in, uh, GA today. Uh, the things that you'll see are in preview, uh, have been described, uh, as, uh, generative actions, which is taking the, those same capabilities and extending it, not just to knowledge, but to But to other systems of record. So effectively, uh, taking an API and putting a generative AI wrapper on top of it so that just like you can kind of do search and retrieval of knowledge, you can also make calls to APIs.

And, uh, convert the results of those APIs to natural language using the Copilot capabilities. Um, that is in preview today, both for the agents as well as directly to the, uh, end user as well for the, for the self service. 

Tom Arbuthnot: So that, help me there. Is that a scenario like, uh, I'm a customer talking to a web chat, but the, the, the, something you build in Copilot Studio could go wrong.

You've authorized me everywhere, but you know who I am, go look in a database and look at my previous orders in a very database API format and spit it back to me in general English. 

Alan Ross: Exactly. So the way I've been describing this is, uh, the highest value that I see is, is taking existing APIs, existing web apps.

So, uh, on the agent side of the, on the agent side, for example, We see agents have to bounce between, say, 10 different screens for any given case. For all the cases they handle, it can be tens of different systems they need to navigate. And especially with the challenges of onboarding new agents, high turnover rate, etc.

The problem is just compounded. So not saying go create a whole bunch of new APIs saying actually the highest value is just take the existing web apps, which our industry has gotten really good about decoupling the front end from the business logic and the database backend, the new front end could be Copilot.

So you could say, Respect the same permissions model, respect the same data model, respect the same business logic, but rather than say, oh, go log into this other system, go traverse this other set of menus, ask the question in natural language, wrap it in that Copilot generative AI wrapper, and you can not only ask questions and answer the question in natural language, but rather ask the question in natural language.

Conduct read operations, but you can conduct write operations as well, rather than say, reset my password, rather than say, run this diagnostic in a web form. You say it in natural language, have the, have the generator, the generative AI and the Copilot be that interface to the same system. Yeah. I 

Tom Arbuthnot: hadn't considered the benefit of that on the, and the relative safety of doing that on the agent side, like the agent has to jump into, uh, you know, The system that has the updates on the shipping and the system that has the knowledge base for how things work.

And the system that has all the, you know, end of life products or whatever it is, if, if, if you can help them and it's a very low risk way, cause they're still, they're still feeding it back to the customer. That's really interesting. 

Alan Ross: Yeah. And, and just some examples of the controls that we've put in place for that.

So, um, you're able to then go drill into the source. So just like people are kind of used to. Generative answers saying, okay, I, I appreciate the answer that you gave me Copilot, but I kind of want to double check. What's the source of this thing? Drill into that before they give the answer. So too, can you do the same thing on generative actions?

So the source on the APIs that you would have called, you can check the source of that to say, okay, I used to log into this other system. Copilot's giving me an answer. Is that the same system that I would have logged in to do that? Oh, yes, it is. Okay, cool. So, uh, it's doing what I would expect before I then take, take the next step.

Um, the other thing that they, that we've done is we've enabled, uh, you to set, like, ask for confirmation for that particular API call. So if you say, I'm not ready to just have Copilot automatically go make this API call. You can, as part of doing the wrapper process, you can say, please confirm before executing.

So Copilot will say, I believe that what you want to do is the following. Are you sure that's what you want to do? The agent has to say, yes, that's what I want to do. And then it'll complete the API call. So we're definitely looking at the realistic expectations of our customers and making sure that we're Listening to their feedback, especially where they're risk averse and they want to sort of control for some of these variables, building that right into the system.

So, um, hopefully that they're not having to build as much of that in ProCode and they're able to rely on our low code, no code SAS, uh, implement. That's cool. And 

Tom Arbuthnot: I guess the customers, this customer system could choose to give. Copilot, read and not write. So it could query, but not execute those kinds of things.

Alan Ross: Yeah. So that's kind of the way that we're seeing this play out in real world. Uh, you're, you're probably right that there's a little bit more appetite to do this on the agent side where, uh, there's at least human in the loop. Um, we're seeing certainly some aggressive customers that want to roll this right out to, uh, their end customers as well.

But. Uh, generally that's, those are the quick wins. It's like, use the functions like summary, which are kind of no regrets, just make an agent's life easier. Uh, then tap into knowledge and kind of use the, as a replacement for the old school, click a bunch of links and try to find an answer. Use it there.

And then from there, you can start thinking about moving from read operations on APIs, all the way to write operations on those APIs. So cool. 

Tom Arbuthnot: And commercials around that helped me understand that. So some of those Copilot scenarios are with the. Dynamics 365 Contact Center License. So I'd assume no consumption rates there, but keep me honest, I'm guessing when you go into Copilot Studio, you're more into consumption and additional costs there.

Alan Ross: Exactly.  So, um, there's capacity packs available, uh, and it's really just a question of the ratio of conversations. So we kind of have a good sense, uh, of how many interactions, uh, an agent is going to have. Uh, once you start exposing it to all customers and self service, Uh, then there's capacity packs, uh, associated with 

Tom Arbuthnot: that.

Awesome. So, I'm getting my grounding now on how it works within Dynamics. How does Dynamics 365 Contact Center relate to, integrate with, work alongside Microsoft Teams? 

Alan Ross: Yeah, so there's a couple of places. Um, first we should note that, um, the underlying infrastructure is shared between Teams and Dynamics 365 Contact Center.

So we rely on the same messaging channels, same infrastructure. Uh, so we get the benefit of, uh, all of that capabilities. Teams is the world's largest, uh, cloud phone system. So we're happy to work closely with the infrastructure that they've 

Tom Arbuthnot: built and proven off. So that's like, IC3 team, ACS, like same That's it.

Alan Ross: You got it. You got it. All right, you're ahead of the game here with throwing those. As soon as we get to Teams, as soon as we get to Teams, 

Tom Arbuthnot: I'm cool again. 

Alan Ross: You're golden. Okay, cool. So, yeah, so we're built on all of that and in fact, just to sort of close off on your question about Uh, other meters. So yes, we do rely on, uh, on acquisition from the customer of an ACS, uh, subscription and the burn rate for minutes, whether they're going to use, uh, direct offer or, uh, direct routing.

Um, either way, it still comes in through, uh, ACS. So that's, that is just like Copilot Studio is, uh, a separate purchase. So, so too is ACS. And again, it's basically. Because of the volatile nature of individual customers and how much, uh, they, how much capacity that they would, they would need there. Um, we have also built, uh, kind of hero scenarios within the integration.

So, uh, the ability to do swarming is kind of one of those key capabilities. Um, where the agent may want to consult subject matter experts that aren't using the CKAS interface. They're, they're just people within the firm, uh, that are using Teams. So we have the ability to swarm a case by bringing them into the contact center interface to troubleshoot.

Make sure that, um, they're a part of the case. And then the agent kind of continues to do the closure of the case. And so what does that look 

Tom Arbuthnot: like on the, uh, on the company worker, knowledge worker side? I can understand that logic in the Contact Center agent experience. What does it look like for the Teams user?

Alan Ross: Yeah. So for the Teams user, they're just getting the inbound message. It looks like any other Teams message. Um, there's clues to let them know that they're, Uh, part of, uh, this, this case. And, uh, we have the ability to kind of send them, uh, additional data on what the context is. Uh, but ultimately it feels to them just like any other, uh, Teams message, most important.

So 

Tom Arbuthnot: like in, in the Teams world, we have obviously the native call queues and auto attendant. We've got a lot of ecosystem vendors doing contact center that kind of, embed into the Teams UI, but this is not that. This is separate experience, dedicated agent experience. 

Alan Ross: Correct. So, um, so the, the interface that we're using, this isn't like go do a screen pop of a Teams app.

It's not like that. We're actually, uh, injecting the message as a regular message into Teams. Makes sense. 

Tom Arbuthnot: Cool. And you touched on, when we talk about ACS there, you touched on bringing Telephony, Voice 2. 0, Dimensions Drive, Contact Center. So, uh, uh, take us through how that works. That's one of the most confusing things, I think, coming from the outside world.

Cause it's, you have to start understanding ACS and Azure Direct Routing. So maybe you can take us through that. Yeah, 

Alan Ross: I mean, any customer has a choice of whether they want to use direct offer, meaning they want to buy phone numbers from Microsoft, purchase those. Uh, through their Azure subscription through ACS, um, or if they want to bring their own carrier and they have, uh, Verizon, AT& T, BT, uh, any of the partners that we support, if they have those phone numbers and they want to use direct routing, they're able to do that.

Keep the contracts with their existing, uh, telco provider, uh, but then have us be the end point. You can kind of think of it as the entry point of being the SBC, Session Border Controller. Uh, that's where it picks it up, but you still need an ACS, uh, subscription to do the, to do the. Awesome. And for 

Tom Arbuthnot: the, for the Teams audience, cause the terminology is similar here, we have.

Teams Direct Routing. This is Azure Direct Routing, which fundamentally is very similar connecting your or a hosted SBC to the cloud, but they are different and parallel. So if you've configured Direct Routing, you can figure it into ACS potentially. That's true. I know there's a number of carriers as well that will do this for you, set it all up as a cloud service as well.

Um, and then whether you're using. Direct offer minutes from Microsoft or you're using, uh, you're bringing your own carrier. There's still a consumption cost there for kind of routing the minutes essentially. 

Alan Ross: Correct. And it's going to be lower for direct routing because you're relying on more of the weight being from the telco.

Makes 

Tom Arbuthnot: sense. And then while we're talking about modalities, I jump straight to PSDN there, but Is the Dynamics 365 Contact Center omni channel as in video, web RTC, web based video, Teams calls? How does that all work? 

Alan Ross: Yeah. So, um, if somebody is on a voice call and, uh, they want to go to video, then we do have the Teams embedded control, uh, inside of that experience.

So they can kind of You know, upgrade to the video screen sharing interface if they want. Um, they can also, uh, route the conversation, uh, to a Teams user as well. So we can kind of move it over to full Teams experience. Um, if that's what they want to do, um, we, we have some, some level of flexibility there.

Tom Arbuthnot: Awesome. So I'd love, I'd love to understand better, I don't know whether you've got any Customer scenarios or Satya mentioned in the earnings call. There's a lot of transition to contact center internally at Microsoft as well. Kind of, I'd love, I'd love to better understand some of those use cases. 

Alan Ross: Yeah, absolutely.

So, um, So Microsoft, uh, support is of course, one of the largest support organizations in the world, supporting basically every kind of language that we do business, which, uh, leaves very few. Um, and, uh, that's our customer zero for building. Uh, this, this tech, our engineers sit right next to their engineers.

In fact, we recently organizationally brought ourselves even closer. Uh, so we're going to continue using one of the largest contact centers as our uh, customer zero for all of our requirements. Um, five years ago or so, they were really working with a hodgepodge of many different systems through acquisitions.

And just like every one of our. Uh, large customers that I speak with all has a similar story. Um, they started going down the path of consolidating down to, uh, Dynamics as their, uh, contact center, uh, primary system of record. And so, uh, they've been able to drive a lot of benefits from, from doing that, as you can imagine, just on the kind of vendor consolidation alone.

Uh, and then we've really continued down that path on especially our Copilot investment. So, as our way of just making sure that we are, uh, really investing in the places that have the highest business impact. Uh, being able to have that relationship where, The Copilot PMs and engineers are able to really sit side by side with the actual support agents and understand their flow of work and build the features according to what's going to have the most impact there.

That's just gone a really long way. Uh, we've, uh, released a set of, uh, metrics that have a rigorous study behind them. So full A B testing control. Groups, Flipping Copilot On, Flipping Copilot Off, uh, happy to send that link out. Oh, please do. We'll put it in the show notes. Yeah, that'd be 

Tom Arbuthnot: great. 

Alan Ross: Yeah, great.

Um, so happy to, happy to, to, to share that. Um, but it really shows the impact of having worked directly with, uh, Uh, our, our Microsoft's own support organization in order to, to build this set of features and, and to prioritize our work. 

Tom Arbuthnot: Awesome. And what about external customers? 

Alan Ross: Yeah, so, um, specifically on, on Contact Center, uh, part of the earnings release that we just had, uh, Lenovo, 1 800 Flowers, Synoptek, uh, all customers that we've been working really closely with.

We have a much. Much longer list. Yeah, that, that were there I'm sure we'll be releasing. And in fact, that's a major portion of what my team is, is focused on is just getting them also across the line and adding to that, adding to that list. So we've got plenty of proof points for, uh, our contact center being, being used to scale, um, thing I love about the one hundred flowers example is the, the Phone number is the name of the company.

So yeah, it gives you an example of, it gives you an example of, uh, of a great, great customer, uh, for customers. 

Tom Arbuthnot: What would you say is the sweet spot for customer type, customer profile, seat count, what like, who's a good fit, would you say? 

Alan Ross: Yeah. I mean, we have designed this thing to support the largest contact centers on the planet.

So. There is no contact center too big for us. Um, we, uh, we've seen sort of the, the largest ones. We're working with some, uh, some of the largest contact centers, uh, on the planet right now. And, and we are, uh, on track to be able to keep up with whatever The paces that they want to on board, um, part of the investments that we've done over the last couple of years is to rebuild our voice stack from the ground up.

So fully cloud native, taking into taking full advantage of the fact that our contact center engineers sit side by side with our Azure engineers. And so it is truly cloud native and built from the ground up, bringing in the nuance, uh, pieces. Uh, with that as well for some of the advanced voice recognition, noise suppression, all of those advanced features, uh, which, which come with it.

Awesome. 

Tom Arbuthnot: What about, what about the other way? So like, like, is there a sensible minimum seat count or minimum scale? 

Alan Ross: Uh, I've been surprised at the frequency of this question. In fact, uh, there is no minimum seat count. So you can go buy a single, a single seat. Uh, we see that especially being a pattern for customers that want to do a pilot with a handful of licenses and they want to see what the real thing looks like.

Just grab a couple of seats and give it a, give it a whirl. I mean, that's good to know because that's not, that isn't the 

Tom Arbuthnot: case with all products in this space. A lot of them have a, have a minimum, either a technical or commercial minimum. Yep. That's awesome. Um, anything else you want to, uh, share with the community, Alan, in terms of.

product where you are, like what to do next. 

Alan Ross: No, I mean, I would just say, uh, expect some really exciting announcements over the next six months or so. Uh, we're just going to continue doubling down on, uh, AI, continue doubling down on, uh, all the investments that we need to be the premier contact center solution in the market.

Uh, we're watching really closely to industry trends and really just listening to customers and, and just making sure that. We're not, uh, as the saying goes, skating to where the puck used to be, but where the puck is going, uh, and just as you started off talking about, uh, how bullish you are on, uh, the future of AI in this space, we are as well, and it really is kind of, uh, one of the centerpieces now of the Dynamics portfolio, uh, so we're super excited, uh, can't wait to hear feedback from customers of all, uh, Sizes, Industries, Geos.

Tom Arbuthnot: Awesome. Well, Alan, thanks for taking time out to jump on the pod. Like you've really helped me and hopefully the community understand how it all works, where it all sits, decoded it. So thanks so much for that. 

Alan Ross: Indeed. Likewise. Thanks for having me.