Special Edition: Shopware Community Day w/BOLDEST

Welcome to Commerce Famous, a podcast that celebrates the celebrities of our industry. Join us for a conversation with the taste makers in ecommerce. In their daily lives, these are just normal everyday people like you and me. But to us, they are commerce famous. Hello again.

Welcome to the second special edition of Commerce Famous. I have a couple of friends joining me, live from Shopware Community Day here in Cologne, Germany. First one is Sydney, and the other one is Rami from boldest.ai. Hi. Hi.

Thank you, Jason. Thank you for having us. Is this so this is, I think I'll start with you, Sydney. This is your first podcast. Yes.

First. And your first time to Germany. I'm I'm hitting a lot of firsts this time around. Yeah. Absolutely.

And so we're live at community day. We just recently, our CEO, Sebastian Hamann, finished his keynote. I'll start with you, Sydney. What is your first impressions of the shopper community day as well as some of the things you took away from the the keynote? Yeah.

Honestly, I'm super impressed coming to this beautiful convention center, seeing all these people come together, and, you know, you guys were really engaging throughout the entire demonstration and and seeing what is in store and also, of course, seeing, you know, Pathway up there with Shopware Copilot was amazing. Fantastic. Well, you set me up beautifully. I'll go to Rami. What is Pathway, and why should the world care?

So Boldest is actually an AI first, agency, and we have our product which called Pathway. Yep. And Pathway is, it's conversational commerce, but not in the most traditional way Yep. Where you can just, put the prompt and you get the product. We have our own, algorithm.

It's called PQRST. Stands for price, quality, scope, on timeline. And the entire idea, we take, any question, from the customer, and we plug in our algorithm, and then we run the algorithm, which is based on, the game theory. And then we provide the customer with especially what they need, specifically what they want. And when we provide them with very limited options, they get confidence because it's true that what they are looking for is what they are seeing on the screen.

One other point about Pathway, which is pretty unique, is that it build the page on the fly using agentic to UI technology to the user. Mhmm. So you don't have a preset journey. It's not sequential pages being displayed anymore. It's just the agent is intelligent enough to know what the user is looking for, interject, build the, page to be a guided journey within the concept or within your, storefront without leaving the your, your storefront.

Yeah. Well, let let me unpack that. So Boldest is a relatively new company, a couple years old Yeah. Founded by, Dilip and, I believe, Keith and you. Yeah.

And it is an agency that's basically designed for today's context. Yes. Is that right? Yes. Correct.

And so tell me a little bit more about the why start an agency in this world today. Like, why? So Dileep, Keith, and I were, part of, a very big agency. It was one of the biggest, privately held agencies before. Yep.

And we served our customers for years and years and years. And then I will say, we started to think about AI from the advent of, I would say specifically, the Gen AI into 2022. Mhmm. And we have seen a vision that everything will change. Things will not remain the same anymore.

And we started to just think and put our own vision about what the future might look like and how can we contribute on shaping the future and the industry and not just being reactive. Yeah. Definitely like everybody else, we had to see the progress of technology with changes in on daily basis. But when the the right time came and we felt like the technology, I would say, still changing, still evolving very rapidly. Like, in in the keynote speaking today, it was mentioned nobody can guess what will happen by end of this month.

Right? I I agree 100% about it. But still, we see that, by combining some deterministic AI with Gen AI altogether, we can really change the industry, contribute to the customer experience. Yep. And for us, when we say customer experience, it cannot be separated from the merchants themselves.

The merchants needs to get share of the love Yep. And they need AI to work for them as well. And, this is where we said, okay. We are going to do what we do best, but everything AI first. Love it.

Commerce famous is proudly presented by Shopware, the leading open source e commerce platform for mid market and lower enterprise merchants. More than 50,000 clients already process over $25,000,000,000 in annual GMV through Shopware. Find out more about Shopware and the best value in ecommerce at shopware.com. And what is the okay. So you guys have a great product.

You've integrated into Shopware. Actually, you're one of two beta integrations into Shopware's Copilot to basically extend the the the value of what Shopware has delivered into these unique offerings. And I you described Pathways, pretty well. How do you how is the market responding so far, to the Pathway product? What has been the feedback?

It's very differentiated. So do people understand what it is and how it adds value? It does. Mhmm. Especially when we speak about Pathway from the philosophy of the product itself.

Yeah. If you look on any market any product in the market, we we spoke about discoverability, and we spoke about PQRST, which is, again, based on the game theory algorithms that we own and we we developed. But another part, if you look on the philosophy of the entire product and about everything we do AI, you need to remember a word here called predictability. That's right. And predictability for us means three things.

It means latency. It means cost, and more importantly, consistency and accuracy. So you need to do the same thing 10,000 times, and you will get the expected result you are getting 10,000 times. And we do this through our own way of data, the right data, the right prompt, the right context, and the harness. That's why, it really caught my attention in the keynote this morning when, Shopware focused on this idea that we are in the infrastructure of ecommerce.

So if you have the right infrastructure, okay, with a product like Pathway, we know how to utilize every part of this infrastructure in the right way and in the right timing to get both the merchants and the customers what they need. Yep. Well, you know, the interesting the reason I asked the question around the positioning of the product, is because people generally like to buy things that have been kind of really productized. So you think about a recommendation engine, you think about a chatbot, you think about, you know, these CDPs, you think about how you tie all those things together. Your product is a bit of a category creator.

It's it's a it's differentiated in a lot of ways. Has that been a challenge explaining to people or do you just show them and they get it? At the very beginning, it it was a challenge Yeah. Until we started to show people the modules within Pathway Yeah. That are really required for ecommerce and it's not required than other use cases.

Alright? So for example, very simple example, if you speak with anybody today about fine tuning, for example, fine tuning a model, any model, they would say, yeah, it's not needed anymore. We don't use it, especially for people who are using AI exclusively for cogeneration. Right? But if you look on e commerce and tone of voice, in some very edge cases, you still need to have fine tuning.

And we included, fine tuning module within our application. You don't need to go anywhere. Okay? You don't need to go any to any LLM provider. You can, fine tune some other from within our application.

Yep. This is one simple example out of so many. When we started to show our customers or everybody in the industry, I would say Yeah. The different models that we have added in addition to our algorithm that is 100% done for e commerce, they started to get it because they have seen the benefit. If I don't have this feature, it will be very hard to get exactly what I want.

It will be very hard to be predictable or get an a predictable agent. And this is where we're able to differentiate ourself. Yeah. But you also go outside of just ecommerce. Right?

You guys are able to access information across many different aspects to make ecommerce smarter. Is that a true statement? It's 100% true statement. The entire idea or our premise to the customer, even with a p q r s t or pathway, is to get the user exactly what they want. I'll give you one example, one simple example.

If you have your storefront in a marketplace setup and you have multiple sellers, and you say, for example, I need to buy a gift from, for my daughter, and, unfortunately, I forgot and it's in two days. So you will select, the gift that you would like to gift to your daughter. But since you said two days, we will go outside of the normal ecommerce. We will see the history of delivery of delivery to promise, fulfilling the delivery of promise of every seller. Yep.

And we tell the customer, since you are in a rush and since this product is available for this specific seller, you may consider it because you really need to get, the gift to your daughter on time. Yep. The context. Yeah. So it's outside of ecommerce.

Yep. 100%. Yep. But for us, as we just spoke a little bit, it's where how to get the right information from the right, system. Put everything in the context so we can get the right product.

Yep. It's the marketer's dream. Right? It's the it's how do you deliver exactly what they want that's uniquely qualified for who they am in their context. And, yeah, that's the marketer's dream.

Yeah. Yeah. And we also help the customers to understand if I buy this, what would be the value if I I buy something else? So we can actually work on dynamic bundling. Okay?

And it's definitely it will not be obligatory for the customers, but it's it's a nice option to know what works well. And it's not, again, prebuilt. It's not like, you know, go and put configuration if you buy this, you can buy that. Yep. It's based on everything you know about the customer, what event or situations they are in, and what product so far is catching their attention.

Love it. And, Sydney, you're in business development. What has been the feedback of the customers when you talk about this? Are they grokking it? Are they understanding the value?

Are they able to kinda understand what they're unlocking here? Yeah. Absolutely. And, of course, there's always gonna be something unique to their industry, and so we're able to kind of say, hey. It's not just for, you know, for telco, for example.

We can actually tailor this, pilot it for you, and show you the value. And so that's been a really, impressive conversation with us. I I love the idea of not telling but showing. I think that's the most powerful use case. Talk a little bit, Rami, about data sovereignty, data control.

What happens if people use your service? Can they opt in or out opt out of sharing all this information? One One of the biggest concerns around people giving their information into these LMs is it gets, you know, shared by their competitors as an example. So how does that concept around data sovereignty play into your strategy? I I believe this is one very, very good question.

So from day one, when we designed our platform in general Yep. We design platform with absolute control. Specifically, there is internal controls Yep. Where you pick the right data only that you would like to share. Yep.

And then you will take another step where what you are going to share with LLM. And this is pretty important. So I would say we have been pioneer on implementing guardrails. And when it comes to guardrails, for example, if I I need to really mention here's PII because it's the most important Yeah. There is no way that PII from our system will go to an LLM.

You know? We have so many strategies. Our customer, while designing the agent, will pick one of them. And it takes the the customer, the designer of the agent, a lot of effort to say, I would like to bypass the rule, and, for example, only phone number be passed to the LLM. So we have it controlled.

The default is it's never passed to the LLM. But, also, again, we are making sure that when speaker control, it's internal for their, own data. And if the customer, by mistake, provide any information, it will never go to Zalenam. Yep. Well, I tell you, one of the things I love about your company, your product, your strategy is you're designed for today and you're not bringing back any of the history of any baggage, I guess, I would say, of of a classic agency having to reinvent themselves, you started with a clean sheet of paper.

And so, I think you look a lot more like the agency of the future, which is, yes, you have a services arm, AI first, but you also have a software product that can help customers grow their revenue. Obviously, you started the business three and a half, four years ago. Was that the design point to be both a services company led by AI, but also a software company? Yeah. It's actually we really believe that, in in the AI, you need to have two things.

Yep. One is product Yep. And domain expertise. But you cannot have both of them separate. You need to have the mix between both of them.

Yeah. So our our core offering definitely is a product. But if you look at it, you cannot really have the product and without providing services Yeah. Because otherwise, it might not be easy to integrate it within your overall business process. And it might not meet their full potential of why they bought the product.

Yes. Exactly. Yes. And one other part which I believe is very important to mention is that we have also another product called PestFinder Yeah. Which is is the same AgenTek design, the same rules, is actually applied, but it's, it serve more than e commerce.

It serve business process. For example, if you have you want to optimize your, business process from the back end, any back end process Yep. You can use Pathfinder. Got it. So both of them kind of complement each other, but one when you have the back end process and the other one is, again, for, ecommerce.

Driving conversions and mass dynamic personalization. Yes. Exactly. Got it. So one of the things we really enjoy about working with you guys, we've been a partner for, I think, a year and change, got a bunch of implementations in flight.

You guys love working on these big complex use cases. One of the components of a complex use case that you guys solve is the ability to take a marketplace technology like a marketplacer along with a shopware, along with the pathway product, combine it to deliver what is next generation experiences. Take that wherever you want. Yeah. I I like again.

You're asking very good questions, Jason. This is very good. One of, of the early use cases that we implemented is related to sellers management in the marketplace, setup in general. And the idea is when you have, multiple sellers on your storefront, what you need to do is really, one, is to make sure that you closely monitor their performance Yep. But then you need to action based on their performance.

Most of the marketplaces comes with, prepacked actions. But what we have really seen and and have been doing marketplaces for years and years, and I would say our team really, implemented the most complex marketplaces worldwide. You know? Yeah. So we found that when it comes to seller management, it cannot be, it cannot be just rebuilt.

Every case by itself is an edge case with, with marketplaces, and it requires a lot of effort and actions. It's a business of exceptions. Yeah. Exactly. So our one of the earlier use cases is how to migrate 100% the storefront shopware and any marketplace, platform together in one agent where you can access from from anywhere.

You can create your own back end portal, or you can use Microsoft Teams or whatever, where you can just manage everything. And with one, either click of a button because everything for us is, presented with text and a call to action, UI. Uh-huh. It might not be available in Microsoft Teams because Microsoft Teams, limitation, not ours. Yes.

But in general, in in how in a click of a button or just, two words or three words in Microsoft Team, you can get everything done into the two systems, and you just rest assured on everything done correctly. And, definitely, I have just to mention that since we are going to maintain data in systems, human in the loop becomes for us a mandatory. The user needs to know what will happen, needs to approve, and when approved, we, the agent actually take care of the actions. Yep. I really do appreciate how you guys are incredibly customer centric and trying to solve real business problems, and software is just an a a a how.

It's more about the problem, and then these are gonna be the LEGO pieces that you assemble to solve business problems. Is that kind of the mindset at Boldest and how you think about this stuff? Beautifully cut. And I love the LEGO reference as well. Yeah.

Fantastic. Well, I'll let you get back to the conference. Thank you for being part of our community, and thank you for coming to Cologne, Germany, to celebrate you guys being one of the pilot members of our, co pilot product. Oh, thank you for having us. And it was really amazing journey with your team, on integrating Pathway with, Copilot.

Yep. Thanks to Lip and Keith. You're missing a great time. I'm gonna thank you. Thanks so much.

It's a nice thing.