Welcome to the Age of "Supervised Automation"

This Retail Technology Spotlight series podcast is brought to you by the Omnitalk retail Podcast Network.

Hello everyone, I am Chris Walton, your host for today's interview, an interview in which we are going to explore the question of how far has AI already gone in running retail operations and what happens to the people who used to run those operations they trust when they.

See their data, their process and AI recommending or doing the same things that their best employees doing.

Each workflow that is built can be self corrective, self improving.

It's also being able to do work that you weren't ever able to do before, which is the crux of what we're going to talk about today.

I myself, I replaced my role within the company about one and half year ago in terms of resourcing management.

Really some work is just too hard to get done by humans.

AI isn't waiting for a press release moment to take over retail and it's already running pieces of promotional planning, merchandising, and perhaps most interestingly, the software development behind the platforms retailers use.

Every day we talk a lot about AI copilots and assistants in the industry, but not nearly enough about the moment teams stop doing the work and start overseeing it, that's a much bigger shift.

And I got news for everybody, it's already underway.

So to help us explore this very topic and, and the question at hand, I am pleased to introduce Anton Spirico, the founder and executive chairman of ScandalWeb.

Anton, thanks for joining us.

I'm really excited to talk to you about this topic.

Hello Chris, thanks for inviting.

Really glad to talk about this topic.

That's what it's really at the edge and where we see some real breakthroughs are happening with retailers.

Yeah, it's really mind blowing.

It's really transformational too.

But you know, before we get into the meat of the discussion, which I know there's going to be a lot of meat on the bone for this conversation today, Tell our audience a little bit about yourself because I was doing, I was doing some background research on you and your, your, your background is really interesting and I think it's, it'll give the audience a good perspective on who it is and why you're an expert on this subject.

Yeah, thank you.

I started as a company 23 years ago and we still run it with two co founders.

It's still privately held.

We started with a small web studio back in Latvia.

Now we are around 500 people working globally and naturally, well, our focus is E commerce and commerce.

And over these 20 plus years we have seen a lot of interesting shifts in how people were shopping from desktop, from mobile, where platforms were deployed, data center on premise, cloud, now AI, naturally.

And that's the most multifaceted, multi age, because it's both a channel where retailers now sell and being discovered.

It's also a tool because as you just mentioned, you can create a lot of software with AI that we definitely will speak later about.

And at the same time it's something that can do the work.

So it's at least three in one and definitely there are more use cases of it.

So this transformation is so far the most exciting where I, I'm very glad we got that way.

This experience, this vision to meet this.

Yeah, exactly.

It's technology.

It's not a tool, it's not a channel, It's a technology, just like electricity.

So the applications of it are yet to come.

That's really good.

The comparison of electricity is really interesting too as we get into this, because, you know, it is, is going to potentially be that as impactful as the invention of electricity or probably even air travel in a lot of way.

As we use further analogies.

I can't believe you started this company 23 years ago.

For those watching on video, you look way too young for that, my friend.

That's one thing I'll say.

I wish I looked as good as you after running a company for 23 years.

23 Years, yeah.

Some years are stressful, some enjoyable, but looking back, it's quite a journey.

But I think AI brought a certain renaissance to us because it's, it's a change of such a magnitude that we really feel that we have reset, that there is a certain mental renaissance in our company.

Mental renaissance, That's a really good phrase too.

I'm curious, Anton, when did that mental renaissance start in the company?

Did it start when ChatGPT introduced itself to the world or were you all looking at this already before that?

How has the company approached that?

Yeah, well, we had some minor experiments with machine learning.

Specifically, some 10 years ago, we built a tool to track sentiment of our customers.

So it came with one painful event when customer unfortunately left citing bad service.

So we wanted to know what's happening before red flags appear.

So we integrated it in our service desk communication and then it helped a lot because teams could see real time satisfaction of their customers.

So that's a great example.

When something that has not been measurable becomes miserable.

Real time, then it, it has its impact.

And I think our kind of massive mental renaissance came when OpenAI released their API so we got not just the chat as such, but we got access to state of the art LLM.

It's funny what you just said there too, because the other part I think about is too, it's like it's, it's also being able to do work that you weren't, weren't ever able to do before.

Which is the crux of what we're going to talk about today as well.

So like, I'm curious, how far has, in your 20 years of running this business, like where are we now in terms of how far has AI actually gone into running the operational layers that drive retail today?

Like how would you answer that question?

Where are we in terms of the stages of it, of AI actually operating things?

Well, I think what we have seen two years ago that AI has been assisting people.

Now we see people supervising or judging, pretty much making a judgment across whether this suggestion is good or not.

So AI has been assisting, now AI is producing outputs and people.

Well, yeah, people are supervised.

So it's kind of supervised automation as we call it with our customers.

So we do not aim especially working with large organizations, Fortune 500 and alikes.

Automation needs to have two, three, maybe four steps before, before that starting from okay, data infrastructure maturity and then eventually it should be supervised automation.

And then when the number of incidents, when humans have been correcting AI drops well, then it can turn into fully autonomous process.

We see some parts of well defined workflows being in this quite advanced stage of supervised automation.

It could be personalization at scale, it could be pricing logic, it could be analysis of large volumes of customer data, certain segmentation options, product discovery through byproduct of LLMs.

They have so called vectorized representation of data.

So the search can be much better or it can become really conversational.

So there are many, I would say focused applications, focused segments of workflow where things turn from being AI assisted to autonomous under supervision.

And of course AI in software development, the speed, the velocity is, it changes, as we say, the economics of which problem now you can solve.

Every merchant we have seen multi million, multi billion organizations, everybody has spreadsheets, all right, Old large spreadsheets, they could never tag them.

It felt like we need three business analytics.

We need this, we need that, we need and then it will take half a year.

Now you can iterate already in a week you can have a solution.

You can iterate 2, 3, 4, 5, 6 times.

In a matter of weeks you will have something that would eventually replace the spreadsheets.

It would give observability of Change what you are doing there, better access control and eventually it will become part of your workflow rather than an isolated document.

Just as an example.

Yeah, but AI in software development definitely brings like a magic wand to a capable E commerce director.

Because the challenge usually is seeing where you need to make a spell.

Right?

Because some people are just like let's make it all autonomous and then somehow.

But if they really the what we usually call normalized problem, the problems they got used to, it's not a problem anymore, just how life is there.

But this exactly the areas that are very beneficial for change, once we change them, the impact is usually exponential.

So basically if I rewind what you said, you basically are saying that you're definitely seeing the application in the software development process within retailers.

You're saying at a 30,000 foot view, you know, as as much as people want to go to full automation, they're still really in the realm of like trying to implement what you said, supervised automation, which I think is another great phrase.

You're throwing out a bunch of great phrases for us already on this podcast.

But they're going to a realm of supervised automation.

So I'm curious then if you, if you, if you go into the retail organizations themselves, are there areas where the software is being deployed to specifically help with things like is it promotions, pricing, like what areas are you seeing retailers start to take this supervised automation approach the most?

I see two kind of mutually distinct areas.

So one is it will look like the customer is being at the top, customer experience being at the top and then certain surfaces customer is interacting like a storefront and then backend what, what is underneath.

Customer doesn't see all the earpiece, warehouse management system, transport management system.

So we usually see that a success comes from going deep into the customer.

So personalization at scale, as I put it for our customers, that it's the first time in a history where millions of your customers can get a brand ambassador that works like your best ever best employee and is available 247 and serves just one person and it works.

We had few projects and I mean we work in commerce so they are transactional.

So personalization plus messenger gave a customer extent to conversions.

Like the conversions increased 10 times comparable to traditional means.

Yeah.

So that's so powerful.

And another kind of opposite side of of the same stack where we see AI is helping a lot is modernizing old legacy systems.

Some of them are 30 years old, IBM or as 400 as they call it initially.

We modernize old systems giving them let's say capabilities that are necessary for this real time cross channel decision making.

So that's where AI helps a lot.

We, if it's.

Well, at some point I can share about the platform we have built specifically to address that.

I'd actually love to talk about that.

Before I do though, I want to make sure like I'm understanding that too is because one of the things I've been thinking a lot about lately is like, you know, one of the beauties of AI is like you said, there's all these legacy systems.

But in some ways if I'm, and I want you to correct me if I'm wrong here, but does a, I'm asking you a question about it.

Does AI now enable us to improve upon those legacy systems more quickly than we would have had the ability to otherwise in terms of trying to rip them out and replace them and do all the things that we couldn't do before?

But does AI enable us to actually kind of find ways to solve that problem of having to get away from legacy systems?

Yeah, absolutely.

Absolutely, yes.

Absolutely, yes.

So I would say it's a radically new venue now that AI offers for large businesses where we, we, we call it erp.

Change is a hard surgery on the business so that sometimes we do, but it's right, right now we work on one replacement that customer wanted to do for 16 years.

So this is not a decision making pace you would expect now in, in our times, right to wait for something 16 years.

But what happens now?

What, what is this, what is this new radical venue for.

For the merchants, businesses having all the RPS that are pain or a multi million project to replace and big risk for business continuity.

So what you can do now is to, to build an operational layer sitting above or in between these systems.

So with AI, well, I mean it's technically was possible before because in the end that's just a code and well, okay, certain AI affordances.

But before it would be as we have seen it would be a multi year, multi, multi million dollar project where there's a chance that stakeholder who started it would already quit or would leave.

Right?

So now it's again it's this magic wand.

But you need to be clear with your wish or a gene in the bottle.

You need to be clear with your wish or what, what data you want to make real time, what system to connect in one say operational dashboard, not just informational but operational.

And that's where we have seen an opportunity to create something reusable because certain rules, they stay the same single sign on certain data Protection, guardrails.

We call it Opera layer because it focuses specifically on operations.

As a commerce agency we, we see need for backends to support e commerce operations, ecommerce operations.

They need to be real time, they need to be comfortable, convenient.

Every update should be timely.

All promise on price or stock delivery should be fulfilled.

Right.

So backend systems need to support this type of promise that we are doing on the storefront.

And it's usually not the case.

But with AI, you can envelope them, you can wrap them into operational layer that would actually connect systems that were never meant to be connected or would make certain decision making layers on top of the systems that are used to be a books of record.

Got it.

And so Opera layer, why'd you call it that?

I think, I mean, I think I get it.

But is it that, is it that blatantly obvious?

It's one of the colleagues who was the first to do this project, he, he somehow came from the obvious operational layer and a romantic root of Opera or the idea that you are.

It is like something that opera is usually something where you orchestrate many voices into something beautiful, at least to certain taste.

And that's perfectly resonated with the idea of what he's doing.

It's a name that.

I like the name and it's a name that makes a lot of sense too.

Because if, if I hear what you're saying, you're.

You're basically saying when we get right down to it, that you know, that, that the advent of AI is, it's not, it's not causing a situation where we need to replace or rip out the ERP system that a retailer is using.

It's really about giving the retailers or the teams that are using it decision making service that was never built in a way to provide that, to provide information or data that they could now get on their own.

Is that the gist or am I missing something?

Yeah, exactly, exactly, exactly.

You are right on point.

I can give you one example.

Yeah, just one example, maybe a few more.

But there is a reason Opera is modular.

So there is one module built for a B2B retailer.

So they have around 1 million products and thousand categories.

So they have a lot of category managers.

They are sending purchase orders to their suppliers.

So they are distributor, right.

So B2B distributor and then.

And it's handled in ERP.

And then they have warehouse management system and then incoming shipments are recorded there.

How's it cool.

Like shipping notes, invoices were recorded somewhere else in the accounting system because they need to Be handled, paid or compliant with some local regulations.

So what has been happening is that at no point in time they had real understanding on what of their purchase orders are actually honored, completed or partially completed.

They did not really reconcile where what is being shipped and received is actually what they asked for in the right quantity, in the right attributes.

Because it has been across three or four systems and only at some reconciliation points like once a month.

They were getting somewhat delayed perspective on that.

Well, naturally with computer vision, with image recognition, we could build a system that scans all the data, that gets the data from ERP, from vms, reconciles it together, validates, verifies calls for human intervention if necessary and gives a real time perspective as well as flagging every, let's say over deliveries or under deliveries that are actually urgent.

So they can react fast on that, not just okay with three, four weeks delay they can say oh actually these guys didn't deliver so we cannot ship.

So they finally for the first time they got real time picture of probably thousands of deliveries per month.

So not only in of course without any change to the erp, VMS or accounting practices.

So the OPERA layer just connects to all of them.

Some of the systems don't have API, but there are many ways how we still can source data in a compliant, comfortable manner.

Not only it gave them the decision maker, the decision making ease and velocity they never had.

But while we always talk about customer experience, there is also, let's call it enterprise experience of how people of what is experience of people working in an organization.

We see many times that brands advocate for certain, let's say high standards of at workspace, but the systems are antiquated.

What do we see when, when a brand, when an organization gives such level of support, of operational support, it changes culture.

So it becomes a cultural artifact and a true, not only commitment, but real material step forward for the modernization of processes that positively impacts culture.

There is definitely ROI on that, but there is also some soft non tangible benefits that I believe are very important.

Yeah, that's one of the, that's a benefit of AI deployments in organizations.

I never thought about that actually could increase the attachment of the organization or the desire of your people to work there because you're making their lives simpler every day versus them having to beat their heads against the wall to get the answers they're looking for.

All right, well, so let's segue then then Anton, because you know, I teased it at the beginning and this is actually probably the conversation that I personally am most excited about.

So let's let's say you have an organization that's, that's taking the opera layer approach.

And you know, let's say they're, they're, they're, they're, they're doing that and they're kind of working on this idea of supervised automation like you said right now, what actually is happening at the team level that used to do that work by hand or used to do that work of checking things over on a monthly basis like you just described, what's actually happening at the organizational level.

Yeah.

In terms of, at a human level, at a personal level.

What we see also in our organization, in our finance department for example, is that there used to be employees and customer organizations that were very valued because they could match three sources together fast.

Right.

Three spreadsheets, export something from collaboration software, match it, map, extract, make a pivot table like spreadsheet stars.

So this skill is, Is not in a demand anymore.

As soon as you automated it with AI, well, AI would help to map all the data together and eventually build the software.

Well, there can be some, also AI decision making on some corner cases, edge cases.

But yeah, this type of skill is not necessary because AI is doing it real time every day 24 7.

These people usually, at least what we have seen in the organizations we have been working, they have more than one skill.

So these people usually stay to improve the system or eventually to envision how it can expand.

Because the road to AI native organization is not a one module.

So whenever there is a capable employee who helps to bring AI to improve it, to flag all, let's say cases that deviate and eventually, well supervise it and eventually let it go.

Well, these employees are usually very demanded.

To bring the same mindset to some other areas that is repetitive takes time and possibly costs money.

So Anton, I'm curious because I was having this conversation with somebody yesterday too.

So let's say when the systems get in place and it does become more about supervision, are we going to see retailers gravitate towards more like general specialists, People that are more skilled in many different things than the traditional specialists because they're going to be managing, you know, so many agents down the line or how do you think about that question?

It's interesting question because just recently we were meeting with some like super old time customers and we are talking about E commerce directors specifically or E Commerce.

Yeah, that once we started these guys were like small entrepreneurs in the organization because everything was new.

Nobody knew how to service retail.

What is order management system?

What are your statuses?

Also Payment gateways were not there, shipping was not there.

There were no all this nice modern solutions that have now.

So they were inventing them.

So they were kind of very, very capable generalists.

Yeah.

And eventually they had to specialize or actually to build their team who would specialize in customer data platform and personalization in organic traffic acquisition, paid traffic acquisition, customer experience experts, A B testers, conversion rate optimization and so on.

So there was so ample opportunities to go in depth and each, each vector would actually bring you return on investment.

Yeah, possibly.

What what we see now, some of our faster moving customers, they again become generalists.

They do they.

Okay, yeah, they become generalists.

They really dip their toys in all kind of waters across the whole organization.

And they usually become.

Well, there are some exceptions, but they usually are the AI champions who actually bring certain vision of how AI can be deployed and where in each of specialized areas of commerce.

But it takes a generalist.

It takes a generalist and the courage to actually disrupt and unbox some processes and actually pick one that would be beneficial for automation.

Wow, interesting.

Got it.

That's why I love this job.

Like I have one conversation with somebody yesterday and then I bring in the conversation today and it actually ends up hitting right on the money in terms of like a nugget to leave the audience with.

So I'm curious, is that when you, when you look at retailers that you're working with in terms of who's handling this transition well, is that one of the big separating factors?

Is it the only separating factor or are there others that are handling it well versus maybe those that are struggling in terms of figuring out how to get on board with an AI transition?

Well, I think for us as practitioners there is a big was a big let's will we.

We see it as a range, as an axis.

So we always kind of nudge customer towards the end of the axis where there is a well defined piece of a workflow, measurable, specific, we can Change something in 3, 4, 5, 6 weeks.

The projects like AI, all this AI transformation does not happen or takes years without any tangible results.

Is where the ambition is to transform something that lacks granularity at all.

Okay, so it's good to have a vision.

Let's build an AI AI native organization.

We ourselves have the same vision, but we are transforming one process at a time.

For example, we have calls just like we have with you.

We have a transcript analysis app that helps our business developers, key account managers, to improve, let's say their persuasion or improve the structure of their pitch.

If It's a new customer.

So we can analyze them real time, give them feedback, give them some, let's say coaching.

Ask for.

Yeah, you would improve Next.

It's a small piece, but it's very practical, very tangible.

The same, the same for organizations strategies.

It's great to have a large long term strategy, but where we have seen millions of euros spent or dollars and already years of work accumulated without any practical application.

Or there are cases where organizations actually build something.

But it's something.

I have a perfect example of few organizations that built it.

They call it let's build corporate brain.

They build it, they sometimes even use some open source LLM, train it, whatever takes GPU effort, time, effort.

But nobody uses it.

Yeah.

Because it was not part of their workflow.

Nobody needs really a brain.

They just want some specific thing being done.

They want to understand.

Okay, I now address John in this organization.

What is the tone of voice of John specifically AI.

Please help me to write this speech for John specifically.

He's CIO and we already have like email exchange, hundred emails.

So please pick up how John really approaches the offer, what he rejects, what he likes, how he reasons.

Yeah.

So something very practical and people even don't want to ask, they want to get it.

So that's the, that's the best.

That's the best candidate for success is when we just give something to people that they use.

We always go from an output.

We don't want a tool like a brain.

Yeah.

We need output.

So we reverse engineer.

So here is a success.

Let's take one step back what happened before.

Let's look here.

Can we improve it with AI?

Okay.

If we can, we can.

If not, then we move further backwards and we find a place where we can generate an output that matters.

And then organization learns, then organization trusts and then they are on a path to becoming AI.

Native organization.

It takes an unusual skill.

Right.

As you guys electricity.

Yeah.

Nobody knew what to do.

As our favorite example is that some guys in manufacturing, they just change their lamps, right.

To electrical lamps from gas.

It was not even like a candle, it was gas already.

So they changed some other guys did what?

They built machines.

Right, Right, Right.

Yeah.

It's why it goes back to the whole con.

The whole.

The motif we were talking about before too, which is, you know, chances are the general practitioner that understands the wide swath of the applicability of AI is probably going to have a better chance of success with this.

I would think inside an organization, if that's your mindset in terms of thinking about you know, how do you apply it?

In what spots and which spots overlap and where can it be used in a new way too?

This skill is still, I think, not very well defined.

What we see, these are people who challenge things and they can spot normalized pain.

Right.

Because otherwise you just improve what you're already doing.

Like, it's like sending emails faster.

Right?

So, okay, let's.

It's easy.

You already have content.

It's easy.

But look at something that is hard, something that your team has been avoiding.

We have a cheat sheet to help our customers to narrow these processes down.

Like something, something that it happened, like that you delegate something and it does not happen.

People avoid it.

For example, one thing, right?

Nobody owns it.

Yeah, nobody owns it.

So these are the blind spot of organizations that are true targets for you that when AI is there.

Yeah.

Then it's.

Another cheat code I can share is that we have built many great solutions that customer did not even consider to be a problem.

They were just complaining about it.

So one way is just to listen what people complain about because they don't think it's fixable.

They just complain.

They think it's just impossible.

Let me complain about it over the dinner or a beer.

And then we tell what if it would be a problem?

How what would be your ask?

And then there's big resistance.

But if they formulate it, then we usually can formulate a solution.

Yeah.

And that's a theme I've heard in multiple conversations this year too.

And it goes back to what you said in the first, in your answer the first question too, which is like, some work is just too hard to get done by humans.

And we as organizations don't always have a good understanding of what that work is.

And probably what the work is that they're complaining or not talking about is a good litmus test for where that, where those opportunities lie.

All right, so let's, let's shift gears a little bit now.

So, you know, we've been talking, we've been talking about, for the most part, you know, the supervised automation concept, but where does this all go next?

You know, how close are we actually to automated commerce?

Like, what's your take on that?

Like, truly automated commerce, like, is that a reality?

Is it someplace we can get to?

How long is it going to take?

Let's just have a discussion on that one, Antonio.

I think the vision of automated commerce is where let's say order is placed, order is fulfilled, whatever package label is printed, maybe robotic warehouse has taken it packaged, slipped and sent it So I think that's definitely something that we can almost see nowadays, but with a big exception that still there are humans who are supervising the process.

Yep.

Each workflow that is built can be self corrective, self improving and, and people now don't need also BAs or team leads.

They, they can just narrate what they need and a system would build a small application for them.

Right.

So if they find some repeatable issue, they can build an application for that, they can build an exception for that.

But still there are humans because either as a system needs to be so perfect and envision what will happen in retail like in two years and it's kind of self adapt and I, I, I don't yet envision such system or there should be still people who are just doing different jobs.

So instead of being between systems or copy pasting or doing as we tell detective work around some shipments and orders and delays, they are overseeing the process, they are judging, let's say AI decision making.

AI feels okay, it's a, it is exception.

So this order is flagged because there possibly stop conflict and then suggest some action and then humans can decide okay, we accept it or we override it with something else.

But I see it's a very exciting process where you see work happening.

You don't need to create certain outputs, outputs are being done, but you need to orchestrate it.

You need to, you need to have this creativity to see what's next.

Because one thing is of course find normalized pain and address it.

But another thing is like with this personalization at scale is to imagine something that humanly was not possible before or was possible in a super small family business, in a small village where you know everybody and you serve them well.

But now any brand, even with like hundred millions, half billion consumers can actually ensure the same level of service.

Oh God.

All right, well, before we let you go, you know, we, you've seen, we've seen a lot of, you know, tech promises come and go over the last, you know, 20 years particularly, or maybe even 30 years if you go back to starter.com.

What's the biggest lesson you've learned in trying to get retailers to actually trust AI enough to hand over basically its operational layer to it.

As we've just been discussing.

What's the biggest lesson you've learned along that front?

Well, as we tell our customers are very smart people because they run large businesses.

So their trust is one of the battlefield.

Yeah, right.

They're smart people.

Trust their hubris maybe.

Yeah, right.

Yeah.

They are not trusting Your slide decks, they are not trusting even your enthusiasm because you can be just blinded by some nice technology.

Would it be whatever composable commerce or headless or something there?

Yeah.

They trust when they see their data, their process and AI recommending or doing the same things that their best employees doing.

And of course when exceptions happen because decision makers they are involved, it's not when everything is great, they're involved.

They usually involve when things.

Then some exceptions occur, some corner cases, exceptions and then they need to intervene to handle them.

So when they see this becomes less than it used to be and they don't need even their mental capacity to address them to help, then they say oh wow.

I myself, I replaced my role within the company about one and half year ago in terms of, of say resourcing management.

Really?

Yeah.

So I, it was super simple.

I just collected say around half thousand or maybe up to thousand of my decisions on certain situations.

I could extract like a decision making metrics and then whenever there was a case, AI would apply all my archive and this decision making my metrics and would explain what Anton's would do and what is a suggestion.

So and then a human would actually decide, but this human would always have Anton by the side.

And when I saw them I thought my God, this is so much better.

Because I'm sometimes in a rush.

Why forgot about something.

But this guy never forgets.

He never in a rush.

He has so much capacity.

He's got access to your entire brain too.

Not the things that you've, you have in your brain but you forgot about.

Wow, that's.

Yeah.

And it's, it's again it's a very small like stratified set of what I'm doing.

So one role, one process at a time.

That's in our opinion is a like a one workflow at a time.

That's a way to autonomous retail.

Yeah, the, the keywords I would take from what you just said too is like it's going to be the easy, easiest.

It's going to be easiest for the retail leaders to understand what's tangible to them and what they can actually feel in terms of making their life lives easier.

At the end of the day too, that's where you're going to want to focus first.

Which makes sense when you get right down to it.

All right, man.

Well that was great man.

God, I could, this, I could talk to you for hours.

This is really, really good, really really great content and it was a really great discussion.

So if people want to get in touch with you Learn more about Scanda Web 2 the Opera Layer.

What's the best way for them to do that?

Well, they can find me on LinkedIn as Anton and Scandi Web, or they can just go to scandiweb.com, fill in the contact form, and they tell that they have seen me with.

Talking with Chris on Omni Podcast.

So that's easy.

We are founders.

We are still very much engaged with the business, so would be very glad to continue this conversation with some business.

Right.

This is really heady stuff we got into and a lot of.

A lot of stuff that's coming to me at the top of my head, too.

And you answer those questions with, with, with great articulation and a lot of insight in ways that I never have even thought about before.

And so, and, and, and also kudos to you.

No one's ever said, like, yeah, reach out and tell us.

Tell them.

Tell us you heard about us.

You heard about us from your conversation with Chris on the podcast.

That's awesome.

So, yeah.

All right, well, that wraps us up.

Thanks to Anton.

Thanks, Anton, for joining us today.

This podcast was produced, of course, with the help of Ella Sirjord, as they always are.

And on behalf of all of us here at Omnitok Retail, as always, be careful out there.