How Williams-Sonoma Inc, Indeed & Heathrow Airport are Improving Customer Experience with Agentic AI

Transcript of a YouTube panel (moderated by Jason) featuring Samir Hassan (Williams‑Sonoma Inc.), Linda West (Indeed), and Peter Burns (Heathrow Airport) discussing deployment of agentic AI (AgentForce/Halle) to improve customer experience and scale operations. Participants describe motivations for adopting agentic solutions, concrete use cases (interior‑design agents, passenger assistant Halle, enterprise agent deployment strategies), early operational impacts (faster response times, higher average order values, workload shifting) and principles for integrating agents into existing workflows and platforms (Salesforce ecosystem).
Chapters
Thank you so much. I'm thrilled to be here today to welcome some incredible customers who are at the forefront of the agentiq revolution. And they're here today to showcase some of the real world value that they're already finding with it. We've got some amazing brands represented here today. Williams Sonoma Inc. Specialty retailer. It sells high quality kitchenware, appliances and home furnishings under beloved brands like Williams Sonoma, Pottery Barn in West Elm. We've got indeed here. It's the world's leading job marketplace. It helps more than 3.3 million employers connect with the right candidates. And we've got Heathrow Airport, one of the world's busiest and most iconic airports that welcomes more than 83 million passengers every year. So before we start, I'd love for you all to just introduce yourself really quickly. Sameer, why don't you start?
Sure. Thanks for having me, Jason. So I'm Samir Hassan. I'm the Chief Technology and Digital Officer at Williams Sonoma Inc. Great.
Linda.
Hello. I'm Linda West, VP of Business Automation for Indeed.
And Pete.
Hello, everyone. I'm Peter Burns. I'm the Director of Marketing and Digital at Heathrow Airport.
Wonderful. We're so grateful that you could join us today. Really great to meet you. Let's just dive right in. I'd love to start with a question for you, Linda. I think a lot of businesses right now are finding themselves at kind of an inflection point regarding AI investment. And how did you decide that the time was right to jump in and get started with AgentForce?
Our objective is to accelerate existing value output within our business. Our core mission is to help people get jobs and to connect those 3.3 million employers with job seekers and do that quickly. And so seeing a platform that enables us to bring that to life inside of an ecosystem that we've already spent quite a bit of time and investment building existing workflows that already do that. So applying AI is really an accelerator on top of that. So I think for us the decision was really around, okay, let's commit to AI and do it in a context where we're already heavily invested in producing the outcomes that are driving our business forward.
And did you have to choose between a bunch of different vendors? Was it a difficult decision?
Yeah, I won't say it was necessarily a difficult decision because we have so much of the rich context that's already available inside of the Salesforce ecosystem. Everything from workflows to data to, you know, human engagement in that platform. All of those things already existed. So I Think it was quite natural to say, okay, we're going to add agentic solutions inside of that ecosystem. But, but I will also say, yes, like, we do ultimately have to make choices between how are we going to build those agentic solutions? Is it on the Salesforce platform or other platforms that, that are, that are emerging on an ongoing. But I think for us, having. Having that connected ecosystem inside of Salesforce is really what drove. Drove our investment here. And we can then, then come in and accelerate what is already good versus starting from the ground up, which is a really different way of approaching this.
Right, right. So you've already got a few, You've already got the infrastructure in place and you're leveraging your earlier investments. So you're not starting from a, from a, from a standstill. Pete, I want to talk to you a little bit about your agent, Halle. Am I pronouncing that correctly? Halle?
That's right.
All right, good. And what strikes me about Halley, is this is an agent to help travelers while they're at Heathrow, and this is over 80 million passengers each year. So how do you build an agent that is robust enough to. To serve that many travelers at scale?
Yes, sure. Thanks, Jason. Well, yeah, we're really excited to have launched Hali. And in terms of context, I'm sure everybody has been through an airport, hopefully Heathrow, a few times. And I think whether you travel occasionally or a regular business traveler, we all want an easy, a seamless journey. And it's worth mentioning we have two objectives when it comes to experience which agent force supports. For some people, traveling can be really stressful or uncertain. So we want them to have completely remove that worry and have a really easy journey. And then secondly, we want people to enjoy their time at the airport. You know, that might be a family break. That's. Well. Or they might be arriving in the UK for the very first time. So we have this phrase in the business. We often say your journey starts at Heathrow because there's so much to offer. So that's the context of experience. And then in terms of agent force and how we scale that, we think about it and have graded experiences both back of house and front of house. So back of house, agent force, and Salesforce powers our customer center. We've got passengers contacting us for many different reasons, mostly through digital channels. So AgentForce helps answer and manage those conversations for us, and that's enabled us to give a better experience and also answer those conversations much more quickly, 40% more quickly for our customers. And that's really important if they're in a hurry going through their journeys front of house. We've launched Hali as a passenger facing agent that answers the questions we all want during our journey. At present they're more information based, but as we mature capability, Halle will be managing more of your bookings, more of our products and services at Heathrow. And, and it's interesting, we have 80 million, over 80 million passengers. And much as we would want to individually hold all of their hands through Heathrow, practically that's not possible. So what Hali and Agent Force allows us to do is effectively scale personalized customer experience. And the thing I like about Agent Force, Heathrow, like many businesses in the background, is a really complex ecosystem of technology and partners and we do a lot of work bringing that together. But what AgentForce does is connect all of that backend into a really simple front end interface. It's convenient and importantly, it's easy to use and human like. So fundamentally that takes the complexity out and fundamentally then helps achieve our objectives of either reducing worry or allowing people to enjoy their time at the airport.
Can you give me an example of what a personalized experience might look like? You know, I can imagine like where's the bathroom or something would be a very easy case, but it sounds like you're pulling together a lot of infrastructure to, to inform these sort of interactions. So what can me an example of what that might look like?
Yeah, sure. So Halle pulls in a whole range of different data sources. The ones that people find most useful are the real time data sources and we have those in different areas like our app and our website. But putting them into one simple place that people can interact with both before, during and potentially after their journey is really key. So for example, live security queue times, we have live immigration wait times, we have live mapping and beacons all around the airport to help you navigate through the airport like you might do on your phone or in your car. And again, if anything changes might be your gate or any other information, you can be the first to know about that. So it's information based at the moment, but using real time information. Going forward, we're building out more mature capability. So an example might be, hopefully Jason, you're sat on the beach on holiday and you decided to stay an extra couple of days. You could simply speak to Halle to get her to extend your parking booking and it would automatically do that for you. Link through to the information we know about you and our payments platform and you can go back and have another cocktail.
Awesome. That's Great, thank you. So moving on to Williams Sonoma. Williams Sonoma obviously is a beloved brand and its stores are very white glove, high touch, high quality. Can you talk a little bit about how Agent force can make those customer interactions feel authentic and bespoke?
Yeah. Jason, I'll give you a compelling example of a set of use cases we're working on. It's all around interior design and Williamson and Mahome, Pottery Barn, West Elm. A big part of what we do is helping customers with interior design projects. I'm going through this myself, just bought a new house and one thing that we know is that interior design is difficult and mistakes are expensive. You get the wrong size, you get the wrong layout. It's an incredibly challenging and difficult process to go through. And this is why William Sonoma Inc. Although Williams Brands offer free interior design services, but with agentic AI, what we've been able to develop and what we are in the process of developing are interior design agents. So these are agents that understand your style preferences, understand your space, understand your home, understand your product preferences, your brand affinities, and are able to help you design and walk with you through the entire process of not just choosing the right product, but making sure it's designed in the right way and that you don't make those same interior design mistakes that people, people can make when they're embarking on these journeys. Whether it's getting the wrong size of sofa, not having enough clearance from the door, having a rug size that doesn't quite fit the space that you're, you're designing into. And so we feel like with AI it's allowing us to deliver that same amazing guided service, but at a scale and level of quality that's just unprecedented.
That's amazing. And can you talk about how that's driving top line business growth? Are you starting to see the impact of that?
Yeah, 100%. So a big portion of what we do business wise is around home furnishing, home redecoration, large scale projects with, with agentic AI and with Agentforce and the experiences that we're delivering, there's just a level of scale there that is, that is unprecedented. And so, you know, customers that might otherwise think, okay, great, I'm going to go, you know, get my sofa from West Elm and I'm done and maybe I'll go elsewhere for my other projects. If you're able to deliver that end to end white glove guided personalized service. That same customer that might have just decided to buy a sofa, buy a single item and think of you as sort of a niche retailer might end up doing their entire space. There might be, you know, decorative accessories and other elements and wall art and other things that they're going to end up purchasing with you. And we're seeing this is, we're really driving Sheriff Wallet. We're driving, you know, average order values. So it's not just about even converting the customers in the same way that they have been. It's that once you can guide them through that experience and you can deliver it with a level of personalization that they're just not used to, they end up purchasing a lot more than they would have otherwise.
That's great. Thank you. Linda, I want to go back to something you said. You mentioned, the challenges and opportunities in deploying agents across the entire enterprise. When we talk about the agentic enterprise, we talk about not just elevating humans, but elevating entire businesses. So how do you bring all that work together and how do you weave this impact across an entire organization? How do you lay that foundation?
Yeah, that's a really good question. And our, our approach has been pretty straightforward, which is this. Like we, we have approached this from the standpoint of what understanding really deeply the work that our, our human teams are already doing and saying, okay, let's break that work into jobs to be done and, and understand what agents might be better at first. First of all, can they do it faster? Can they do it more efficiently in service of our mission, get jobs? So where can we speed up that process of an employer connecting with the right talent? Where can we speed up that process of a job seeker searching for an opportunity? And if you look at that through the lens of what our human teams are already doing, you can say, okay, what can agents do faster or better? And then secondarily, what work that humans are better suited to do can be accelerated by the support and insights that can be provided by AI. So that perspective of just starting from the value we're already driving as a business and as a company to help people get jobs, breaking down the work from the perspective of human teams and then bucketing it into how do we drive this work via agents so that humans can be freed up to do more revenue generating activity? And then how do you accelerate the tasks that humans should continue leading with the support of AI and if you look at it through that lens, you end up not kind of throwing spaghetti at the wall and just trialing and testing things. It's truly an approach of how do we accelerate existing outcomes that we're targeting that are, that we know, are providing value in the world via agents. Because really managing a fleet of agents is the same sort of context as managing human teams. Obviously the way that they're, the way that you manage them is slightly different, but the work is similar. And so if you look at it through that lens, you can, you can really start to drive business value versus having kind of a trial and error type approach.
You know, you mentioned using agents to do kind of what machines do best and thereby elevating humans to do what they do best. And I'm curious, you know, Samir, Pete, maybe you both can speak to this as well. What are, as agents are taking on more tasks, what are your human teams able to do more of that they weren't able to do before? Pete, why don't you start?
Yes, sure. So for us, Agent Force, Halley, it's very much around improving overall experience. And of course, with an Airport, over 80 million passengers, we have many different service needs, both customer segments and individual needs, which change depending on your journey, time of day, lots of different factors. So we've always tried to deliver what you might call a blended service, helping both those who want low touch service as well as those who prefer higher touch and have higher touch requirements. If I use that back of house and front of house analogy, back of house in our customer center, feedback from the team using Agent Force is they genuinely love it as it helps them manage lots more conversations, it's quicker, helps them with case wrap ups, leaves them to focus on other activity. In 95% of cases, we've not had to make any manual amends. Agent Force is doing that for us and that allows them to do a number of different tasks, including, for example, manage our VIP terminal where many of our bookings come through our customer center. That is more of a high touch experience and it allows us to give a better and deeper experience because we have more time. So that's back of house and then of course, front of house. Hallie is there answering passengers queries. She sits alongside our brilliant internal colleagues and they provide their own blend of customer experience. And so for us, it's about giving them the time to help passengers, especially those with more complex needs. So for example, as I mentioned earlier, if you quickly want to know the live security queue times, you're in a rush, you can get that really instantly through Agent Force. But if you have, for example, specific mobility needs when traveling, then the team have more time to support you fully through your journey and overall, you're more satisfied and get a better experience.
Great. Samir. Do you have anything to add?
Sure. You know, I think when we started this journey, there was a mindset around, this is an efficiency play, which you take the repetitive task, you automate them, and that frees up our people to do what they do best. But what we're actually finding is that it's not this sort of divide and conquer approach. Yes, we're. Yes, we're seeing the efficiency gains, and yes, we are automating the repetitive tasks, but AI is actually amplifying that creative and strategic work that our people are doing. So it's not just about freeing up their time, but when you equip brilliant minds with this kind of technology, we are actually seeing better strategic and creative outcomes from folks, whether it's product development or in supply chain or with merchandising, or in photography and visualization or in technology. And so it's really both. Yes, it is about freeing up your people, but it's also about amplifying your people. And we're seeing amazing results when the tech is activated in the right ways.
Wonderful. That's excellent. Well, thank you all so much for your time and your thoughts. It was really great speaking with you. I'm sorry we're out of time. Hope we can continue the conversation again at some point. Thanks for showing us how you're embracing Agent Force across your organizations, and we're looking forward to seeing what you all do next. Fiona, back over to you.










