Microsoft Partner WinWire All In On Generative AI, Copilot

‘Every workshop that we talk to, you plan for 50 people but 100 people are signing up for it. … So this is a hot topic for everyone. They need partners like us, the team that I’m bringing in, to go deeper into technology to scale and to speak to the customers. And we are able to provide that,’ WinWire CTO Vineet Arora tells CRN.

Vineet Arora and his team at Microsoft partner WinWire Technologies are all in on the vendor’s generative artificial intelligence push.

The CTO of WinWire Technologies—a Santa Clara, Calif.-based 2023 Microsoft Partner of the Year finalist and No. 233 on CRN’s 2023 Solution Provider 500—has transitioned his role to focus on technology from Microsoft and Microsoft-backed OpenAI that can quickly create content, analyses and summaries based on user prompts.

“The scale of the investment that we have planned is [incomparable] to anything else that we have done in the last 16 years,” Arora told CRN. “My entire time is going into this. … The opportunity size, the technology, speed at which it is evolving, requires 100 percent—110 percent—dedication from our side. So the scale and the level of focus has been different from anything else that we’ve done in the past.”

[RELATED: Microsoft Partners: This Is Your Copilot Speaking]

WinWire Invests In Microsoft AI

As part of the transition, Arora heads a small team of about 12 WinWire employees—data scientists, data architects, data engineers found through an internal hackathon earlier in the year—focused on partnering with Redmond, Wash.-based Microsoft to provide this technology to customers. WinWire has more than 1,000 employees worldwide.

Since April, he’s conducted a seven-city roadshow with Microsoft technologists to demonstrate what generative AI can do in health care. He’s done more than 24 envisioning sessions with customers around generative AI, and the company is already piloting generative AI solutions for health-care and life sciences customers. All of this before some of Microsoft’s most impressive generative AI tools become generally available.

And he even publishes two to three articles a day to a Teams group chat called “All Things OpenAI.”

Generative AI promises to be a major boon to Microsoft, rival vendors and their solution providers.

A recent report from investment firm Wedbush Securities said that more than 50 percent of the Microsoft installed base will adopt this AI functionality for enterprise and commercial use over the next three years.

“We view the initial pricing details as very bullish for the total addressable cloud AI market opportunity for Microsoft that could increase cloud revenue annually by 20 percent by 2025 based on our estimate,” according to the Wedbush report.

The firm predicts an $800 billion AI spending wave over the next decade and that in 2024 AI could comprise up to 10 percent of overall IT budgets compared with about 1 percent in 2023.

Read on for more of Arora’s views on generative AI and where WinWire is headed.

Why is WinWire positioned well in this race to generative AI?

We have been in business for over 16-plus years now. I was at Microsoft for many years. When I left Microsoft and started WinWire along with our CEO Ashu Goel, one of our agendas was always to be aligned very closely with where Microsoft’s direction is.

From the very early days of Azure and everything else, we have made sure that we are always part of the best and the latest and greatest that Microsoft is investing their time on.

So I’ve been to [Microsoft’s headquarters in] Redmond, [Wash.], worked with the product team and everyone else every year when they choose the next set of bets that they work upon. … When Microsoft started … looking at all this technology, it was natural for us to look at all of those exciting opportunities that they will offer to a partner like us.

So we are what is called a managed partner. What that means is we get very closely connected with the field team. … We took a strategic decision as we were coming back from the holidays. We will pivot my role, at least for the foreseeable future, to only focus on OpenAI and everything OpenAI. ... All our projects, all our customers are only Microsoft. There have been skeptical statements made and skeptics who have been like, ‘Why are you just putting all your eggs in one basket?’

And Microsoft, the way it has increased its presence in everything in the last five, 10 years, and in the last five or six months on this technology, I think our bet has been paying off.

Why invest so heavily in Microsoft generative AI and Copilots?

I’ve been doing a lot of things on data and AI and app dev, mobility, cloud—everything, for many years. But January onward, me and a small team that I formed is entirely focused on where we will need to partner with Microsoft in providing generative AI solutions.

So in that, what we have done is … one, we of course have our internal center of excellence, where we have access to all the services that many customers even don’t have access to because we are a managed partner on the Azure Cloud.

I have been a part of connecting with the product team and understanding what all is coming down the pipeline. … Second, because we partner closely with the field [team], we have already been doing a lot of actual customer-facing road shows and workshops.

So we just finished a seven-city road show that we did with the health-care group at Microsoft where we showed our examples, our offerings in the area of OpenAI and generative AI. My team and I presented that. … What we have worked on as an organization is, of course, build capabilities.

Build demonstrable solutions. Build a lot of offerings. … We actually have packaged offerings around advisory, around envisioning and even solution accelerators that are built on top of generative AI technologies already out there for many weeks that customers can leverage [and] Microsoft can promote to their customers. … This, we believe, is both an external focused area, but also we are a software development firm.

So it’s also very much internally focused in an area for operational improvement in our entire SDLC [software development life cycle] that we are looking at. … We are doing some proofs of concept already with customers. Pilots that we’re doing with HLS customers—health-care and life sciences customers. And large customers like Adobe and others.

And we are also implementing these technologies in our own software development for test cases, documentation generation and code bugs and identification of those. … Microsoft is a big channel for us. Probably the biggest channel for us.

So all the workshops and road shows that we’re doing, we are the key partners. If you do a search, many other people are now waking up. … But we have already been doing this for three months. I did the first workshop at the end of April, in Bellevue. Imagine that. So we have already been on the forefront of that.

At the same time … our existing customers that we have tons of business [with] already, they are also looking at how to leverage this technology.

So in parallel I have been doing those sessions also to expand with those customers so that somebody else does not need a new partner to come in and say, ‘We are an AI expert.’

We [WinWire] are the AI expert for you for a long time—data and AI. And we are also generative AI experts. … Both our commitment to Microsoft as a platform, Microsoft as a partnership and to our existing customers, it is creating a lot of buzz out there for us as a company.

How big of an investment is this for WinWire?

Every year we align with Microsoft. But the scale of the investment that we have planned is [incomparable] to anything else that we have done in the last 16 years.

So when mobility came into the picture. When IoT came into the picture. When cloud, of course … as data and AI itself came into the picture. We have what we call an incubation of the practice.

That’s what I spend time on. And we get a couple of people to get to know more, to get trained and then build the practice over time.

But here, I already have a dozen people dedicated to this. So you can imagine, the scale that we are looking at is very different.

My entire time is going into this compared to in the past. Take this as—25 percent of your time, something new has come in, but you have other things to also work upon.

So I think the opportunity size, the technology, speed at which it is evolving, requires 100 percent—110 percent – dedication from our side. So the scale and the level of focus has been different from anything else that we’ve done in the past.

What do you need Microsoft to do to help make generative AI more channel-friendly and enterprise-friendly?

The biggest one right now is customers did not plan to invest into a technology like this. … If Microsoft has invested so much in OpenAI as a platform, how much are they ready to invest into incentives for engaging partners like us? … We are hoping for an opening of the wallet and getting incentive programs. … which customers are also looking forward to. … As a managed partner, I have access to lots of preview versions of technology that many customers even have never seen yet.

But when I am able to get on to a call, I’m able to show them that. Having that edge for us, I would expect Microsoft to continue to provide to us so that we can continue to evangelize them. … As I work with their field team, they are hiring a lot of people, they are scaling up their team, but they will always fall short of the demand that is generated out there.

Every workshop that we talk to, you plan for 50 people but 100 people are signing up for it. … So this is a hot topic for everyone. They need partners like us, the team that I’m bringing in, to go deeper into technology to scale and to speak to the customers. And we are able to provide that.

On the technology side and the platform side, there is improvement happening on a daily basis—enhancements, needed features that are coming in. … One of the key things I keep on hearing from customers … content safety … accuracy monitoring, and everything else. And I think that area needs a lot more focus in the next few days and weeks, I would say, for customers to believe that, yes, technology can do a lot of it.

But how do I trust the results? How do I measure the results? What is the interaction that I need in terms of automation versus human involvement, because there is still human validation that will be required toward this step process.

And something that they have already enabled … is working with your own enterprise data. That is a very, very critical thing. And I knew from the very beginning that that is going to differentiate Microsoft, because Microsoft plays in the enterprise space.

And it’s not like the ChatGPT or [Google] Bard where you’re working on internet, publicly available data. Yes, they can do wonders. But what about your legal information? What about your financial information?

Or—we work with life sciences—how about regulated industries information? … I’m actually entirely focusing on that area, streamlining all these generative AI capabilities in a secure manner against your enterprise data, and you’re not just going to ChatGPT and putting your secure information out there. … That is going to require continued investment, continued enhancement in the technology area, giving confidence to the customers. … One of the most interesting comments that I heard… is that, if everybody’s using the same model that OpenAI has released— GPT 4 or GPT 3.5—what’s my differentiation if I were to use that?

The way I explain that is you have to look at that as the baseline. You have to build solutions on top of it. … So OpenAI provides a lot of APIs. … Microsoft is providing a complete platform, like they have provided for the last 50 years. … They provide a platform and you build innovative solutions on top of it.

And things like Copilot are really going to streamline leveraging generative AI in an enterprise scenario. … I’ve been in the Microsoft ecosystem for 35-plus years. … The CoE [Center of Excellence) or the incubation team that I’m building has to have knowledge about what else is happening out there also.

We are focused on how Bard is coming into the picture, how AWS and others are coming into the picture.

Everybody has their own approaches. My current assessment is Microsoft is miles ahead of everyone in terms of bringing it to the market and working with customers where there will be real ROI. And it’s not a consumer toy just to play around with.

What was your ‘aha moment’ with generative AI?

When I started building our own solutions… the time that it saves for no need to actually train the models and train any of these technologies on your data.

I was able to provide … a document or an MSA [master service agreement] document … and ask it to process it in different manners. And I didn’t need to train the system as such.

That capability has never been there in the years that machine learning has existed. And that’s one of the slides that I show is that you have to train the data.

You have to have task-specific models that have to be built. The whole transformer model collapsing all of that into one area. And I can code and I can do text-related processing, all of it into one place in a very simple to use API manner.

The very beginning—I’m talking probably three or four months ago when it started becoming real—is that the same API can deliver different results with small changes in prompts was the aha moment.

The way the Microsoft product team, I think, is thinking about monitoring accuracy, handling hallucination, building what is called prompt flow … that’s just mind-blowing. That’s the development environment where we are going. That prompt engineering is popular. … Having a tool and an environment to be able to work with this prompt in a very professional and in an enterprise manner is the second aha moment lately.

What has helped with this fast, high interest in generative AI?

Customers have a lot of use cases. But I also tell them that this is not a silver bullet. Let’s be very careful about that. .. You have to look at use cases where this has the biggest ROI for the organization.

My approach for Copilot is where you will have to build your solution but add on this as capability to every solution. … And while everybody will experience that in the Microsoft suite products that they use on a daily basis—Teams and Outlook and Word and everything else. I’m talking to everyone, and it’s a little bit tough for them to understand that in the beginning, but when I start showing them examples that as you design those solutions, think of a Copilot, think of a custom product that you’re building, a Copilot with that.

A couple of big product companies that we are working with, they are very much bought into that. That’s the approach to look at it.

And then the ease with which I think customers can get going on this, that has been also very, very helpful.

When IoT came into the picture about 10 years—or whatever years—back, there were sensors that you had to worry about. There was latency and network connectivity you had to worry about. Do we do it on the edge or not? And there were so many different protocols and everything else.

I did a lot in IoT. We as a company did a lot in IoT. But the speed at which customers could adopt that probably took years. Here, we are talking weeks that they can get going on this front. … I have spoken to some customers where the CTO is as excited as me—if not more—getting certain things going on their own overnight with the team and then, I’m telling them, you will need to do it in a secure manner. You will need to do it in a scalable manner.

Look at Azure OpenAI. Not just directly working with OpenAI APIs. That is good for learning and experimentation. But for your production solutions, you should be looking at an Azure-hosted complete solution.

What have generative AI-related engagements with customers looked like so far?

Having done new technology adoption discussions over many years—not just this—one of the things that I focus upon very much early in the game is defining what a typical road map may look like. What a journey for a customer may look like.

Now, there are lots of those journey maps around cloud adoption and everything else. I’m just talking about generative AI-related technology adoption itself.

We are keeping it to ourselves right now because that’s probably an IP [intellectual property]. … But when I speak to someone, that’s the first thing that I show. That you’re not going to jump into building full-fledged production-ready solutions off the bat. That there is a step-by-step approach that you need to take to that.

When I’m speaking to CIOs or CTOs or both of them, our typical engagement starts with what do you see out there, what are you envisioning.

Telling you about what are the capabilities—and yes, please tell me if you have already read enough and you know about it that I can skip some of that content. But there is a lot of confusion out there. … There is confusion and there is apprehension about the technology. ...What is ChatGPT versus GPT? What’s Azure OpenAi versus all of this stuff?

So I demystify some of those things. I typically take a half-hour session for something like that. Show them examples and everything else. Then we envision use cases specific to their industries. … If it’s health care, if it’s in manufacturing, if it’s retail.

Some of those are the biggest domains that we work with. Giving them ideas for what is possible in your industry using this technology. … That’s the breaking point for us. If we can get you to identify one or two, maybe three, use cases that you would like to see this technology be tried out on, that’s the next step. That’s the proof of concept or pilot that we sign up for.

Customers may or may not be willing to invest a lot of money into that. But we are investing our time, Microsoft is investing their money and time into doing that.

Doing a prod pilot or a PoC is the key first step. And then … do I have a strategy to integrate it into my core solutions that I’m using to serve my customers and my customers’ customers? That’s when the Copilot discussion starts to come into the picture.

You are going to have a long-term road map along with your enterprise road map, budget for it within the next fiscal year … budget for services from a partner like us and start somewhere.

Otherwise, I think your competition will certainly adopt it in some manner and you will be left behind. … Not everybody has the time to experiment—we are able to supercharge the adoption of that technology into their areas. … We do what we call a quarterly business review … in every QBR, we talk about, these are the capabilities that we can bring to you.

And then it follows up with the next level of discussions over the next few days after that, that they can incorporate it into their budgeting, their planning and everything else. They are Microsoft’s biggest customers, some of them. And as Microsoft starts bringing those technologies [to market, we are there to help them adopt them in the right manner. And that’s that’s a promise that I provide to them.