Welcome to No Jitter’s Conversations in Collaboration. In this conversation, we spoke with Omar Javaid, the SVP and Chief Product Officer of Avaya. On June 15, 2023, Javaid joined Avaya. Most recently, he was the SVP & GM at Qualcomm, responsible for their global software product portfolio. Prior to Qualcomm, Javaid served as President of Nexmo, Vonage’s API business unit, after starting as their Chief Product Officer. At Vonage, he was part of the executive team that transformed the company from a telco into a UCaaS, CCaaS, and CPaaS enterprise software company, culminating in a $6.2B acquisition by Ericsson.
No Jitter spoke with Javaid after Avaya announced that its current CEO Alan Masarek will retire at year end 2024. That news had been well covered by two of our contributors (Avaya’s Big News and Big Questions for Avaya), so this conversation focused on how Avaya plans to execute on enabling their customers to make the technology choices that best suit their needs, whether that’s on-prem, private cloud, public cloud or hybrid cloud. Some of that will be enabled by reinvestment in Avaya’s existing solutions, partnerships and focusing on where they have expertise.
No Jitter (NJ): It seems like there’s a constant tension within large enterprises – which Avaya serves – where they have a huge number of legacy systems and/or on-premises systems and they’re reluctant to move to the cloud and maybe it’s not beneficial for them. Can you talk a little bit about how Avaya is meeting those needs of these large customers?
Omar Javaid (Javaid): Shortly after I joined, we spent quite a bit of time and effort on ‘voice of the customer’ research to help us understand the customer base. We found that Avaya, like a lot of companies in the space – and I know this will sound like ‘mom and apple pie’ but it's true – has a great customer base. It's a tremendous asset and basically saved the company.
The main thing was we wanted to retain customers, we conducted research, listened to them, and we changed our approach based on that research. We have all those products, but we're not trying to push our customers to the cloud or on-prem or private cloud. We want to do what's right for them and we have a product roadmap and technology investments that reflect that.
[We needed] to understand this customer base because we wanted to retain and grow it. What our customer base was really frustrated about is that vendors – including ourselves previously – were trying to force them into a certain direction. There was…and is…this mega trend to go to cloud , which is real. But if you're a Bank of America or a Southwest Airlines, and you have 30 or 40,000 agents, it's a complex setup, right?
Some of them are your own agents, some of them you outsource to [business process outsourcers] and they're spread out across the world. You may have, as many of our customers do, voice infrastructure that's been there for 30 to 40 years. So while they're technically savvy and have the resources, they're [also saying these] are business critical systems and you vendors should not be forcing us to make decisions that don't make sense for us.
So we looked at [all of] this research and decided we should actually be customer centric. And what that meant for us is we're agnostic. In other words, as Avaya, from a product perspective, we have on-premises products, we have private cloud products, and we have cloud products. And then for the first time in many years we reinvested in our on-premises products to improve those. That's been very well received by our customer base because a large percentage of them have some form of on-prem – they could be on-prem in their own data centers or they could have ‘private cloud,’ which is taking the on-premises and putting it in [Google Cloud Platform] GCP, AWS or Rackspace or something like that.
NJ: Could you talk a little about that – the roadmap and investments? Over the last year or so, you’ve rationalized the product portfolio, announced multiple partnerships and even made an acquisition.
Javaid: [Part of] being agnostic is having products that address their business needs on-prem, private, public – and we have hybrid, as well.
Another thing we learned is that our top 1,500 customers account for the lion's share of our revenue. On average, our top 1,500 customers, have contact centers with 1,800 agents and above. Our top 500 customers have contact centers with 5,000 agents and above. So, if you think about it, our top customers have contact centers that are larger than most publicly traded companies.
Contact centers in general, but especially those in large enterprises, are ecosystems. Avaya recognized this many years ago and another strength of the company are its partnerships and alliances. Take WFM [workforce management], for example. What problem does Verint solve? If you're Southwest Airlines, you have 30,000 agents. WFM is like a specialized HR package to manage a contact center workforce – 30,000 [agents] is a big problem to solve. We have a long-standing partnership with Verint in this regard.
Even if you look at our top 1,500 customers it's difficult to segment them – we have – but in many ways, they defy segmentation. What I mean by that is for some, Verint provides a great solution, and that partnership works well. In others, like healthcare or insurance, there's either some special bit of regulation or they have greater needs for something like outbound calling. That's where these partnerships come in really handy.
And again, [these partnerships] flow from understanding the customer and being customer centric and saying, here's where Avaya will play and continue to play and here's where we’ll partner.
NJ: Can you talk a little about what you’re seeing among your customers with respect to generative AI? Enterprises, especially the large ones have huge concerns data privacy and security, etc. How are they looking at embracing generative AI in the contact center?
Javaid: If you look at a lot of announcements coming out of OpenAI, for example, where they see the early impact of AI in enterprises is, broadly speaking, in customer experience. That's not surprising because customer experience and customer care is, for a lot of these big companies, a significant investment. Anything they can do to optimize and enhance that investment they're going to look at very carefully. I think this is why you see a lot of early use cases and implementations around the customer experience space and it's playing out in several ways.
In the contact center, we’re seeing agent assist, next best action, call wrap up – those types of use cases. And you make a good point about data privacy, data control, and then also local versus cloud models and things like that. I’d say there's a lot of interest, and that interest is translating into pilot deployments. But it's not like they're fundamentally transforming all aspects of business today. It has the potential to do that, but this is early days still. So, there's a lot of legitimate interest that’s translating into real world albeit limited deployments.
Now, what is limiting [many of] those deployments are the pieces you just mentioned – caution around data privacy. You may be aware that the EU released its AI regulations. I confess that I've not read the entire document, but I've read analyses of it. [Unsurprisingly], there's long way to go but I do think AI needs to be regulated and there needs to be some guidance from regulators. The EU has been at the forefront of this and whether or not you agree with how they're thinking of the regulations, set that aside. The U.S. will follow suit and then you have [efforts] in other parts of the world – Japan is looking at their own [regulations], as is China.
The ‘net net’ of that is if you're a large company, and you're global, which is almost all of our customer base, it's a rate limiting step in the near term. They see the potential [of AI] and they're excited about it, but they don't want to go too far too fast and then either run afoul of regulations or have to reverse course. These kinds of customers tend to be cautious. They're doing interesting things, but they're taking it up to a certain point. So that's one.
The second thing is related to that, but there's more to it. They're also being cautious because you have very savvy technology organizations. They've seen many of these technology shifts before. And so they're asking questions like: ‘Okay, OpenAI is a leader today, but will they be the leader in two or three years from now?’ [Their answer is]: ‘We don’t know.’
[Another question is]: ‘These large language models are the rage now, but is that durable?’ [Their answer is]: ‘We think it's durable, but what else is coming?’
NJ: How does Avaya go about providing the capabilities that companies might be looking for with respect to AI?
Javaid: We have a long heritage and history, and I would say, core competence in voice. The internal AI development we've done is around voice. An aspect of our product is called the Media Processing Core where we do AI enhancements to improve voice quality. That's not only relevant for the duration of the call, but in any contact center there's trillions of hours of call recordings – these companies might have 10 years of call recordings – so think of enhancing the audio quality because it's ‘garbage in, garbage out.’ You can then [perform] all sorts of AI and ML techniques on those recordings, such as feeding them into internal models for sentiment analysis. That's where we have focused today in our own R&D.
We do the rest with partners – Google, Cognigy, LivePerson. There we're being aggressive about partnering with multiple companies because the market is moving really fast and we want to provide our customers with the maximum amount of flexibility. Some of our customers may have made a big investment in GCP, so they want to partner with Google on Gemini. So, we’ll have that partnership. The same thing with Microsoft, Amazon and AWS, OpenAI, Anthropic, all of those.
As Avaya, we're not in any position to compete with OpenAI, Anthropic, Google, Amazon or Microsoft. We don't have the balance sheet to invest billions of dollars into a back end to create own set of transformer [models], custom research, our own LLMs or anything like that. Even if we had that kind of money, I don't think that would be the right investment for us.
We have a ton of data and our customers have a lot of data, so let's improve the plumbing. That data can flow in so our customers can then say: ‘Look, we don't want to use OpenAI or Microsoft. We [prefer] an open-source model and we want to put it on our own premise.’
For whatever reason, some of our customers want to do that. We don't want to stand in the way. We'll just say, look, here are the things that you need to do. If you want to take and build or customize your own models, then this is what you need to do. Here's the data, here's the call recording information, etc. Or if you want to feed [your data] into OpenAI, or Microsoft’s version of it, or Google Gemini, etc., here's how you would do that.
(Editor’s note: For more on Avaya’s Media Processing Core, check out this Nvidia-hosted webinar. Also, the call recordings must be transcribed before AI/ML can be used. The better the audio quality, the more likely it is that the transcript will be accurate and thus, the better the AI/ML output – hence the ‘garbage in, garbage out’ comment. Kelly and Kieller touched on this topic in their LLM article series.)
NJ: What would be your message to readers about Avaya’s technology portfolio and roadmap over the next year to 18 months?
Javaid: The main message is we have made, and will continue to make, a very big effort to be customer centric. We have made a renewed investment in on-prem. For those that like on-prem and want to remain on-prem, we have you covered. If it’s private cloud or public cloud, we have you covered. We also have you covered with hybrid cloud – and hybrid is where I think we're doing something unique, which is you can keep your voice on-prem and bring all the ‘cool AI stuff’ in through the cloud.
So, we try to understand you and we hear you. We're here for the journey. We want to make you successful – and we want you to be able to do this with minimal disruption, hence, ‘innovation without disruption.’ I think the roadmap will reflect that. We've already simplified the roadmap, and we'll simplify it further, because we're going to consolidate and bring some of these [new] capabilities into the products.
Editor’s Note: Check out No Jitter’s Avaya news coverage over the last year.
