8 Minutes With Boris Grinshpun: Unified Platforms, AI Myths and What Customers Should Expect

Boris Grinshpun joins 8x8 as Global Vice President, Product Management, at a moment when AI is reshaping customer experience across the industry.

As the head of Product Management, Grinshpun will lead 8x8's global product management organization, driving the product roadmap and execution across the company's integrated portfolio of Unified Communications (UCaaS), Contact Center (CCaaS), and Communications Platform as a Service (CPaaS) solutions. He will also play a central role in translating 8x8's product vision into capabilities, accelerating innovation, and delivering seamless customer and employee experiences.

With more than 20 years of experience building enterprise technology platforms and a strong track record of product innovation we sat down with him to learn more about his philosophy, mindset, and way of working.

You've spent your career in customer engagement. What was the appeal of 8x8?link to this section

Boris: A couple of things. What drew me was how 8x8 brings employees from the front, middle and back office together in a single solution. I was also drawn to the unified platform 8x8 offers, and I believe the emergence of AI will have its greatest impact across the entire portfolio.

The executive team, including Sam Wilson and Hunter Middleton, was a big part of it too. They bring high energy, lots of prior experience and a unified vision for where 8x8 wants to be.

What should customers expect from you?link to this section

Boris: Three things guide how I work. Reliability comes first, because customers run their organizations on these tools. Then a unified platform with features that are useful day to day. Then innovation, AI or otherwise, that’s easy to adopt.

Your career began in financial services, and you've gone on to build products and teams. How does that background shape how you think about customers and what you ask of your product managers?link to this section

Boris: What I learned above all is that a product has to solve a big enough pain point, and customers need to be brought along in the solution.

Too often, technology organizations run out ahead of where customers are mentally and what they're able to adopt. I'm a believer in bringing customers along on the journey: introduce features that deliver value right away, then advance them.

I think of my product managers as having two feet. One is in the landscape of technology and where it's going. The other is rooted in what customers are doing and struggling with today. Marry the two, and act like the CEO of your area with total product ownership.

You've tracked the convergence of UCaaS, CCaaS and CPaaS for years. Why has it taken so long to happen, and what does a truly unified platform unlock for customers that separate products can't?link to this section

Boris: Looking back at UC, it was such a large category that many companies focused on it alone. These markets had enormous potential so companies in each could keep growing in their own lane.

Now it's bleeding over. Today those products are expanding beyond their original categories. In the age of AI, CCaaS platforms are doing the same, because orchestration has become a much more important layer. Conversations are the liquid gold flowing through the veins of every organization, and companies want to get the most out of them to power their own tools and features.

So: vertical solutions, big TAMs, natural expansion. In my view, AI has accelerated all of it, and players like the CRM vendors are starting to see the value.

What's the biggest myth about AI in customer experience?link to this section

Boris: That if I put AI on top of my existing processes, my problems will be solved. Automating inefficient processes or poor customer experiences just makes them go badly faster.

I encourage organizations to rethink their operations from the ground up. Make sure your data sets are valid and holistic. Make sure your knowledge, whether tribal or documented, lives in a known place. Then optimize.

There's a lot of talk about what happens when AI reaches the next level. I think we already have the capability. The LLMs today are advanced enough to automate a number of business processes. The question is how to deploy them intelligently and in the right way. For the foreseeable future there's still an art to knowing when to automate and when to let humans do the work and make the right decisions.

What does the best version of human agents, virtual agents and AI copilots working together look like?link to this section

Boris: There are best practices, especially for rudimentary interactions or predetermined actions with set outcomes. When you need overrides or exception decisions, that's where human labor comes in, and it has to work well alongside the virtual agents.

What I find is that the handoff isn't seamless, because organizations haven't planned for those moments. They also haven't planned for their contact center employees becoming specialists. The entire tech stack is being transformed in front of our eyes to accommodate this.

Despite the best practices, every organization will be different. It's the difference between a chain restaurant and your local place. The recipe each organization cooks up will be its own, and people will experiment a lot to learn what works best and what works least.

How do you decide what not to build? What's your approach to prioritizing a roadmap when everyone, from customers to sales to executives, has a must-have request?link to this section

Boris: The intake problem isn't any different than it was years ago, AI or not. It's always been important to work out what's valuable and what isn't.

What's more important than ever is the go-to-market motion around a feature. When everyone can build everything quickly and copy your features, you have to think about other moats. A product manager now has to ask: Will customers adopt this quickly and easily? Will it be obvious how to discover it? Is it priced appropriately? Is it marketed appropriately? Do we have the right level of support?

Otherwise you end up with AI slop at the product level: too much of everything, all the time, and nothing that solves a customer problem. I encourage our product managers and product leaders to think long and hard about that whole flow.

What does great customer experience look like to you, and where do most businesses still fall short today?link to this section

Boris: We work in the best and most fickle job, because customer expectations are always shifting. If I'd told you years ago that a package would arrive at your house within an hour, you'd have been blown away. Now it's table stakes. What we provide is a constantly evolving continuum.

At the core, though, expectations haven't changed much. Customers want you to understand their problem without over-explaining it or explaining it many times. They want to be known, with their history and the relationship they have with your organization. They want the problem resolved quickly and easily. If you can anticipate it, you're a step ahead. They want clear expectations and timelines, and a solution that meets their needs. They're always looking for a moment of surprise and delight. And they want to be treated fairly, professionally and with empathy.

Your secret sauce is how you stitch together the technology, the people, the virtual agents, and the digital and in-person experiences.

Which CX metrics matter most to you, and which ones do you think are overrated?link to this section

Boris: I'll start with the reverse. Ask a contact center leader and they'll talk about average speed of answer and average handle time. Those were operational metrics that served as a proxy for an environment where you were constrained by how many human agents you could staff. That's changing.

I like first call resolution, although tracking it has been elusive for most organizations. Some are looking at revenue per call or value delivered per call. I like those even more, though they're harder to measure. Customer effort score is also not a bad one.

These metrics will evolve. People are starting to look at resolutions, certainly on the virtual agent side, and it'll be interesting to see how human metrics change as more virtual agents join the mix. I don't have a firm prediction. I just know it'll be changing.

What are the biggest shifts coming for CCaaS and UCaaS over the next few years?link to this section

Boris: A few capabilities are fundamentally shifting. One is personal agents, and how broad adoption across the world could change the way people interact with organizations. Some people are betting on voice-to-voice agents, some on messaging agents, and some on protocols like MCP as the way agents talk to enterprises.

I expect we'll also see more digital twin scenarios. You can already see it happening, but people may have multiple companions doing tasks for them and for their teams. There's my team, and then there's a broader virtual team executing other processes.

Organizations will, I think, also gain real insight into what's happening in their business: where customers have issues, and where there's friction below the surface of the customer experience that causes pain in the outcome.

Finally, I believe the industry will move beyond systems of record platforms, CRM platforms and just communication platforms toward platforms built around the work that needs to be done or resolution platforms.

What's one thing about how businesses buy or use communications and CX tools that you think the industry has gotten wrong?link to this section

Boris: I get the feeling people buy tools because they think the tool will solve the customer experience problem on its own once it's deployed. We in the industry try to build best practices into the tools, but organizations should first assess their current business processes and communication motions, and decide what their ideal state looks like. Very few spend the time up front on that hard work.

From there, think about the solution that fits your desired end state. Don't try to boil the ocean. Take a phased approach: solve the immediate pain points, come back in a later phase to make deeper changes, leave room for innovation, and make sure you can measure outcomes so you can improve.

So: think about your future state, pick a provider, phase it out, iterate, and buy only what you need. I often see organizations buy every bell and whistle and then use a tiny sliver of it. Products are complicated these days. Don't overbuy.

What's a customer problem you've seen up close that the industry needs to solve?link to this section

Boris: The blending of employees in physical locations with contact center employees and digital channels. Today the physical world and the centralized remote channel function completely independently of each other.

Imagine walking into a store where a prior conversation is already synced and in front of the employee helping you. Or the reverse: you've done something in person, and that's acknowledged in the contact center, whether it's a purchase or a complaint. That would be fundamentally different, and it's a fundamentally different way to think about UCaaS and CCaaS.

What's the best piece of product advice you've ever received?link to this section

Boris: There are many, but one comes from a fantastic manager. He'd tell me, “Boris, don't try to be brilliant. Somebody's already solved this problem. Go learn from it.” We sometimes think we're solving novel problems, and mostly we're not. What matters is how we bring the solution into our particular product.

The second is working from first principles. Over-engineered, over-solutioned products are never successful at launch. Customers want to see a product evolve and grow. They're not always ready to unwrap something completely novel right out of the gate.

What's the one trait you won't compromise on in a product manager?link to this section

Boris: Ownership. You can teach and develop many things in a product manager, but you can't teach someone to drive it like they own it. It's the most important trait: feeling full end-to-end ownership, whether that's ideating, marketing the product or even helping out with support.