At Genesys Xperience, Tiffany & Co. took the stage with Miratech to share the story behind one of the fastest global cloud migrations in luxury retail sector.
This is a condensed and lightly edited transcript of the conversation that took place between John Grieco, CTO of Tiffany & Co., and Erik Delorey, Director of Innovation at Miratech, recorded live at Genesys Xperience.
Event Recap: Tiffany & Co.’s Cloud Journey with Miratech
Erik Delorey (Director of Innovation, Miratech):
Good day, my name is Erik Delorey, Director of Innovation at Miratech. I’m very excited to introduce John Grieco, CTO of Tiffany & Co. We’ve been working on a global cloud migration project—large, complex, and fast-moving. We’re dealing with a global contact center operations as a retail base; and for a very premium luxury brand. So tell me about the need for speed in your organization and what brought you to this way forward when it came to your cloud migration from an on-premise system to Genesys Cloud.
John Grieco (CTO, Tiffany & Co.):
Hi everyone, thanks for joining us. First, let me ask—who here has bought something from Tiffany before? (Audience raises hands.) Thank you all that have been by one of our stores or online and bought a piece of jewelry from us. Hopefully, your experience was exquisite.
I tell everyone that every breath I take is about how I enhance that customer experience? That’s all we’re about at Tiffany, white glove, you’re the only one that exists in our world, and we want you to buy more and more, and we want to go on a journey with you throughout your entire life and your family’s life with us.
So, on with the speed here, they always say at Tiffany: “John, what speed do you like going?” I say “faster” because there’s never fast enough, we need to get it done now. LVMH acquired us four years ago. LVMH is a big company, Louis Vuitton, Moët, Hennessy, so we’re a piece of a big landscape of lots of big brands and we’re following in some ways but leading in others, and here is where we’re leading.
We have about 11 global contact centers. They’d been using archaic technology with little investment for years. When I joined three years ago, I was told about this project that had been cut year after year—for 10 years. I said to myself: “Well, year 11, we’re going to get this thing going.”
So, we built the case, justification, did a big RFP, and I’m very happy that we brought Miratech in to help me because one thing we had to do was implement it this year. Many of the vendors that we brought in from an RFP perspective laughed in my face: “John, you’ll never get this done in nine months, it’s impossible, I’ve been in this business twenty something years, and it takes years to get done.” I said, “Oh my God, I don’t have years. I have nine months.”
Miratech said, “We can do it in six to nine months.” And we’re doing it in seven. We’ve done all the Americas and EMEA, and APAC will finish next month. Six months of execution after planning.
The reason we have contracts that are ending at the end of this year is that we’re leveraging new vendors from a telecom perspective, so it’s a big shift into changing what the old-world Tiffany was and now adopting some of the new world, which is we’re leveraging NTT and Genesys with Miratech, so that’s what we’re doing. Seven months, the quicker we get it in, the more I save every day.
Erik:
I appreciate that, John. We took a look at this and we said, okay, these locations have to get migrated, they have to have these basic features, and we’re going to build a roadmap for getting you to the next level. But for a project that is cut every year, what kind of internal stakeholder and community did you have to build to get people to buy in to the fact that not only are you going to actually do it this year, you’re going to do it at a speed that people are telling you is not possible?
John:
To me, it was a no-brainer. Our contact centers are Tiffany’s second largest revenue generator after the landmark store in New York City. If you add up revenue across the 12 or 13 centers, it’s number two.
And yet, our agents were juggling 18 applications during calls—Salesforce, an old Compass application, Cisco Jabber, Teams, and I was, “Oh, my God, this is crazy.” The customer was on the phone on hold while agents hunted for information. I wanted a world where customers are never on hold and agents already know who’s calling, what they’ve bought, and what they are interested in.
We can sell twice as much based on just giving the agents the technology. And the problem was, there was never a voice of the agent. The agents needed better tech, and they needed a voice. I became that voice. I pitched it as doubling profit potential by giving agents the right tools. Step one was getting to a single UI with integrated data flows. That’s what we’re delivering now—with Salesforce, Teams, and more integrations to come.
And more importantly, I needed a vendor that was going to get it done in nine months, right? And I don’t think we were able to find one until, I would say, Miratech answered my RFP.
Erik:
Moving beyond the technology—you built the business case, got executive support, and found a vendor who could deliver. But with a brand like Tiffany, there’s also the need to protect the brand and agent experience while migrating data to the cloud. I imagine security and governance concerns came up—how did you push this across the finish line?
John:
Absolutely. I mean, to me, this wasn’t a technology project, so when I sold it, it was as a change management project. We needed to change the mindset of executives, managers, and agents. I think a lot of folks were comfortable. Now we’re pushing a whole new workflow, a whole new model, a whole new way of doing business but, at the same time, we can’t disrupt the brand, the image.
We have customers calling, so a lot of the managers of these call centers and directors in each of the regions across the globe, I had to really influence them, sometimes nicely, but sometimes with a little New York muscle.
This is why the pressure was there. Miratech needed to do this in nine months, but also train them in this new technology, so we built a great UAT system. We have great project managers within Miratech and Tiffany, working hand-in-hand. And, let me tell you, it’s robotic how we go into each call center. We run a script, which is basically bringing them in, showing them the technology, training them, answering questions, and hyper-care when we go live, sitting with them.
I wanted Miratech on-site for some of these. I mean, I don’t know if they offer that all the time. Maybe they do, but Miratech flew people in, and we had people ready to go on the floor to support one of our biggest call centers, so that was a good partnership that Miratech did that. I sit there in steer-cos, and I try to find an issue. I’m one of those. It can’t be all green. There’s got to be something.
Erik:
We are making sure that the project is handled in a best practice way, especially with Tiffany, because we consider ourselves experienced engineers at Miratech, so we’re about making sure that the experience is good for Tiffany & Co. while, by extension, making sure that the external messages to the customers are consistent and enhanced by our engineering effort.
Now, when you look at the next phase; we’re getting up to level four and level five agentic AI. Where do you see that really assisting the frontline agents with managing customers, and where do you see it maybe directly impacting Tiffany’s customers?
John:
Yeah, absolutely. I definitely want to get to level four, level five. I’m at level three now. I have the bots, they can summarize things well, co-pilot, I got Genesys technology, so I feel we’re about a three with day one implementation.
We already have a list of day two stuff. Our focus will be integrating Genesys with as many upstream and downstream systems as possible. This will equip agents with more information to enhance customer experience and drive sales. Our goal is for Genesys or the bot to begin making decisions on behalf of agents.
I think at some point, there’ll be enough information that it can help the agent pitch: “Based on the information I got, you should go for the Elsa Peretti bead necklace”. We think that would be the next purchase based on their history. Now, it takes some time for a human to figure that out. But at the end of the day, a human is going to sell to a human.
We’ll link bots across the ecosystem: Genesys, our IT bot Jewel, and LVMH’s bot Maya—so an agent can ask one system, and it routes to the right bot behind the scenes, returning a seamless answer. That’s what I want to build.
Erik:
I think we’re well on the pathway to get you there very quickly. I think you’ve got a great partnership with Salesforce, and we’re implementing the new CX Cloud version just after the holiday season to enhance the experience of the agents and the supervisors.
When we take a look at it operationally, and the benefits that this unleashes for the supervisor community, and we’re improving the agent experience, we’re going to increase a lot of processes and efficiencies at the agent level, and lower churn. What does that give the teams an opportunity to do next?
John:
There are a few areas that we want to focus on. One is getting data. I think with the old system, we didn’t have much visibility into what was happening, how productive we were. It was basic technology, that’s how old it was. So now, we’re doing a couple of things.
We’re creating an onboarding bot for our global contact centers. So as peak season hits, we onboard a bunch of customer contact agents. I want a bot training them, teaching them. Right now, we are heavily relying on a train-the-trainer model with leadership stepping in during the holidays. With Genesys and Miratech, onboarding will be automated so agents are productive immediately. That’s why churn concerns me less—because I know new hires can be ready from day one with the bot.
Second, reporting. I’m all over Miratech, they’re building me dashboards, and this is where I love the expertise that we don’t have, because we didn’t have this visibility. And the information’s great. I want to see deep information at the agent level—insights managers can use to make smarter decisions.
Maybe they realize they can shift schedules to improve efficiency. Or they see patterns they never noticed before, just because the data is finally in front of them. That’s the real win: giving people new information that sparks new thinking. Instead of doing things the same way every day, they’ll have the insights to say, “Wow, we could change X, and then implement a better SOP at the frontline.” To me, that’s Nirvana.
Erik:
Yeah. When you look at those frontline SOPs, giving agents the right tools doesn’t just boost productivity—it unlocks their ability to innovate.
For years, it was always, “I wish we could do this,” but then the budget got cut. What we’ve built together has created real excitement, and that excitement is fueling new ideas. Agents are writing the stories themselves, and those stories are going straight into our backlog for prioritization and execution.
AI thrives on feedback, and that’s exactly what’s happening now. Instead of being blocked by “maybe next year,” agents are empowered to say: “Here’s what we need, here’s how we can work better.” That’s the real win: investing in agents—not just self-service technology—so they can upsell, cross-sell, and continuously improve the customer experience.
It’s a really exciting time. And as we start integrating data and enabling bots to talk to each other, I’d love for you to share your vision. You mentioned Jewel earlier—the central repository for how things get done at Tiffany. How do you see Jewel integrating with Genesys to create that seamless, almost intuitive, agent experience.
John:
A couple of things. I’ve been in IT for 30-plus years, and I can probably count on one hand the projects my users actually liked. Usually, it’s complaints: “Another IT upgrade…” But this one is different.
From the start, I could see it in the agents’ eyes. During onboarding, I thought, these agents are miserable with what they have today. So I said, “I’m going to get you what you need. In return, you get me what I need.”
Now, the agents are fired up. They’re testing, they’re asking for more. Our day two list is endless—and I love it. These requests are coming from the frontline. So, we’re creating an agile team, running sprints, delivering improvements every couple of weeks. Some things require us to arm-wrestle Genesys for future releases, but there are plenty Miratech can implement right away.
And we’re tracking results. Just yesterday on the steering committee I asked: how many calls have we handled on the new system? Thousands? Okay—now let’s tie those to sales and compare with thousands of calls a year ago on Cisco Jabber. Adjust for inflation and other variables, and let’s see the difference. To me, success comes down to two things:
One, the agent is happy, loves the tech, and wants to come to work.
Two, the customer loves the experience and buys more.
Now, on bots. I am adamant: I will not give my users 1,000 different bots. No “here’s Copilot, here’s Jewel, here’s Genesys.” That creates madness. We need one simple front door, where the agent asks a question and gets an answer they understand.
Genesys today is mainly for our contact center agents. If they have an IT issue, that goes to Jewel, our IT bot. And then we have Maya, the LVMH bot, which knows about all 80 brands.
My vision is that Genesys talks to Jewel, and both talk to Maya. They share common information. So, if an agent asks Genesys, “What time is the tech bar open in the New York office?” Genesys doesn’t know—but it asks Jewel. Jewel has the information, passes it back, and Genesys delivers the answer.
That’s what I want: all the bots speaking to each other, but the agent only sees one door. Behind that door is every answer they need. Just open the door—that’s the future we’re building.
Erik:
Thanks, John. It’s been a pleasure partnering to protect the Tiffany brand while accelerating modernization.