The AI Employees Running Your Amazon Brand While You Sleep

Steve Hutt:
Welcome back to eCommerce Fastlane. I am your host, Steve Hutt.

Steve Hutt:
Today we’re going to get into what AI employees actually look like from inside a real Amazon business, and whether they live up to the hype. My guest today is Chad Rubin, and this is actually his third time on the show. He’s the founder of a company called Crucial Vacuum, which I think is now called Think Crucial. I believe he’s the co‑founder of a tool called Skubana, which was eventually acquired. I had him on the show back in episode 13 during the Skubana days, and gosh, that was, I think, July.

Steve Hutt:
So it was literally eight years ago in 2018 when I had him on talking about the Skubana tool. Then we had him on again last year, episode 384, with a product that he launched called Profasee — it’s P‑R‑O‑F‑A‑S‑E‑E, Profasee. The original concept was a dynamic pricing tool for Amazon brands, and today he’s here with Profasee Ultra. From my research, he’s not really calling it software. His language is very intentional.

Steve Hutt:
He’s calling it AI employees that actually run your Amazon business while you sleep, and he even has named agents, which are quite interesting, and each one owns a job. We’ll get into pricing, PPC, demand planning, catalog — it’s quite interesting. The numbers are incredible. I just did a quick bit of research: there’s one brand, PF Harris, that added about $215,000 in annualized profit with just 15 SKUs. Chad’s customers have unlocked millions of dollars in incremental profit now.

Steve Hutt:
I’m really excited to jump into this episode. My plan is to unpack what exactly is happening under the hood right now, who this product is built for, and who it isn’t built for. There’s a gigantic commerce shift happening right now, and I think it’s quite interesting even as it relates to Amazon. So, hello Chad, welcome to the show.

Chad Rubin:
Ooh, that was quite the intro. I appreciate it, and thanks for having me back. It’s crazy to think it’s been eight years. I know I was number 13. It’s surreal to say that out loud. You’ve been busy, man.

Steve Hutt:
It’s wild, yeah. And thanks again for taking time to jump on. I really do appreciate it because I think you’re in another phase now. I love the original concept of Profasee because I really felt that the dynamic nature of Amazon — customers and merchants knew there were ways of creating incremental profit based on pricing structures, competitiveness, timing, and all that kind of stuff. It’s interesting how you saw that use case, but then decided to expand into this whole Profasee Ultra and AI employees.

Steve Hutt:
I think it’s a very interesting framing. Can you talk a bit about why it’s different — your concept of AI employees and this Ultra version now versus the AI tools that a lot of people are flogging?

Chad Rubin:
Yeah. Firstly, I want to say the last time we spoke — actually, episode 13 — I was running Skubana, and at that time I was trying to convince ecommerce sellers that the future was multichannel, managing inventory across channels, and everyone thought we were nuts. It turns out we weren’t. Skubana got acquired in 2021. I took a few seconds to breathe and I started Profasee, the dynamic pricing platform that we talked about in the last episode. Then we did a thing: we essentially started building AI agents for Amazon brands. So what happened? What was the flip, what was the switch?

Steve Hutt:
We…

Chad Rubin:
In March, two things happened. Claude Opus 4.7 came out — a new, updated frontier model that was so intelligent and so different from the rest of the models that it could make decisions. Then something else happened, which was the architecture around OpenClaude. That’s when you saw AI agents, or an AI workforce, being able to be built with a structure and architecture around it. When that happened, I was like, “Oh wow, this is my time.” It reminded me of back in the day with Magento. I know I’m dating myself — I’ve been 20 years in ecommerce — but a lot of people made a lot of money on Magento. I felt this was another version of Magento, just an AI version. It’s open source and we can build something around this so that these sellers who are getting squeezed can automate, decide what to keep human, and decide to spend money to get their hours back.

Steve Hutt:
That’s interesting, because your platform now is a multi‑agent platform. It’s interesting that you realized it wasn’t just about pricing alone being the competitive moat — there were lots of other pieces that make a successful Amazon brand.

Chad Rubin:
Yeah, I think it was a couple of things. Number one, our software — the old, legacy, first version of Profasee — that’s just an agent. An agent can actually change pricing for you.

Steve Hutt:
Right.

Chad Rubin:
The switch was, and I think this is going to happen regardless of Profasee, that people are going to move from a software stack to an agent stack, whether it’s ecommerce or a different industry. They’re going to start building agents. Agents will replace their software tools and replace either agencies or humans in the process. So I was like, this is super interesting. It’s a wedge into a much bigger market where the total available market we can capture is very, very large. Pricing is one pain, but there are other pains in the ecommerce space: PPC, for example, and forecasting inventory.

Chad Rubin:
You talk to more and more of these sellers and they’re like, “Hey, I don’t want to do it, I want somebody to do it.” But what if we can have an AI agent do it? The math made sense to focus on this instead of focusing purely on pricing.

Steve Hutt:
Yeah, exactly. And it’s interesting — I love how “cute” it is that these AI agents are actually named. Let’s name them now and talk about what their use cases are. With Claudia, she’s referred to as the chief operating officer. What does Claudia actually do inside Profasee Ultra?

Chad Rubin:
Just like any traditional org chart, as a CEO I don’t talk to everybody on my team. I talk to the people at the core, front and center, at the top of the org chart, and then their work showers down. It’s the same thing with Claudia. Claudia is the ultimate reasoner, observer, analyzer, and delegator. She is the COO. She’s the point of delegation, she’s super smart, and you work directly with her.

Chad Rubin:
Then she goes and instructs the experts to do what they need to do. She comes with the platform. The second AI workforce employee we’ve been working on is Marco. Now you may ask, “Why Marco? That’s an interesting name.”

Steve Hutt:
Yeah.

Chad Rubin:
There’s actually a story behind it. At Profasee — the dynamic pricing platform — we have a PPC team. People want pricing, but over time they want to combine pricing and PPC together to achieve a greater outcome and more profit. If you can control both demand and price together, you have a really interesting mousetrap to make more money. We had this PPC team. It was very hands‑on, very focused on services, which is a very human‑based thing with headcount. Marco was one of our best employees.

Chad Rubin:
Marco decided to leave back in March. When an employee leaves, you lose that institutional knowledge. There’s a trickle‑down effect — it affects everything in the company. I was like, “You know what, Marco, I know you’re going to leave. I want to honor you and I’m going to build an AI agent and model it after you, in your image.” He was honored and a little weirded out at the same time.

Chad Rubin:
Then it became a thing internally. “Hey, this is the way we’re honoring Marco and we’re going to build a brain similar to him.” He was a great guy on the team. Now we’re going to have that institutional knowledge internally at the company. There’s also a second‑order consequence: as these models get better, we are AI‑agnostic. So it doesn’t matter in this nuclear arms race whether it’s Claude, OpenAI, Gemini — they’re all going to get better over time, which is beautiful. We’re going to be the beneficiaries of that. We don’t need to pick a horse. We can choose which model to integrate into our platform — all of them are integrated right now — and as they get better, we just implement that into the platform.

Steve Hutt:
Right. Some of the language, for the techies out there, is that what you’ve built with these agents sounds like “harnesses.” A harness uses the reasoning power of the large language model, but your harness on top allows the execution to actually happen.

Chad Rubin:
Yeah, yeah. We built a lot of this behind the scenes. When you come to the platform, the platform looks beautiful, but there’s a ton happening around: what are the instructions, what is the memory, how does that memory persist, how does it remember today from yesterday? Let’s just say Marco makes a big change yesterday — it has to know it also made a big change the week before, how long the lapse was between those two changes, etc. There have to be guardrails behind it. It’s like child locks and guardrails — things that keep the brand functioning so you can have the best outcome possible.

Steve Hutt:
It sounds like coordination is your competitive moat right now. It seems like if one agent is flagging a stockout or a stockout risk, it would be interesting if another agent automatically pauses ad spend on that ASIN. I’m curious to understand if that’s actually what’s happening, because there are separate tools that can do those sorts of things. But I think your competitive moat now is that these agents can talk to each other and make intelligent decisions based on those outcomes.

Chad Rubin:
Yeah. I’d break that question down to be a little bit more technical. We ingest a ton of data in the background and we do that using traditional software methodology. It’s deterministic — it comes into the background. Agents have the opportunity to read that data, so they can read all of it and deeply understand it. Marco has access to all your inventory data and can read and understand it.

Chad Rubin:
That’s the first layer, and that’s what we do today. We’re still at the early stages of AI — agentic AI — building an AI workforce to replace agencies, tools, and humans at the same time. The second step, the future frontier of this, is that AI agents will be able to have their own stand‑up meetings and cross‑pollinate the changes they’ve made, what they’re seeing, and how they can influence each other’s decision‑making process. We’re not there yet, and we’re going to be there, but that is definitely part two of this coordination.

Steve Hutt:
Very cool. I’m always interested in these early alpha testers — they’re more like beta testers right now — that are using the platform. I brought up the PF Harris example, but we can talk about that one or another brand, even anecdotally, of a direct‑to‑consumer brand that’s on Amazon, running your platform, and is open to trying things. Here’s what their life was without Profasee Ultra, and here’s their life now because of it. Is there anything top of mind that’s fresh, where you can say, “Yeah, I know firsthand what’s going on with this brand in this industry”? We can call them out directly if it’s public; if not, we’ll just talk anecdotally about some successes.

Chad Rubin:
I think right now, at the stage we’re at — this is all very, very early, first‑frontier stuff. If you’re adopting Profasee Ultra, you’re on the cutting edge, the bleeding edge, really, of agentic AI. In the future, this will be much more prevalent for everybody. Right now this is what we’ve been heads‑down on, and there have definitely been early adopters. Today it’s an amplification, augmentation machine. What that means is it’s kind of like Tesla. I don’t know if you’ve ever been in a Tesla with Full Self‑Driving — they call it FSD — but you still have to supervise it, you still have to be behind the wheel. That will change in the next, let’s say, two years, maybe sooner. They just came out with the Robotaxi, which allows you to be in the back, with no steering wheel and no gas pedals. But right now you need somebody behind the scenes who sets the strategy on day one and then delegates that to Claudia.

Chad Rubin:
So for those brands, it’s really amplification. These agents don’t take a lunch break, they don’t take sick days, they’re working 24/7, heads‑down on your business, getting into the nooks and crannies to help you make more money. That’s exactly what they’re doing for the early‑stage customers we have today.

Steve Hutt:
There are brands right now that are exclusively on Amazon, but a lot of people listening today like this multichannel approach and are in a lot of marketplaces, with Amazon being the number one for them. How does Shopify play into this, when your wheelhouse is the Amazon world? The reason I’m phrasing the question this way is because I’m trying to think about the source of truth from a financial perspective. There’s a lot going on in Amazon, but usually there’s middleware or some kind of connection back to Shopify through reporting. They call Shopify the mothership — that’s the core, your owned real estate for your business. What’s your mindset around the Shopify business as it relates to marketplaces like Amazon, and how your platform integrates and works together?

Chad Rubin:
Yeah, so just like Skubana: when we first launched Skubana, we were focused on Amazon, and then we made a bet on multichannel. We understood that Amazon sellers can’t stay only on Amazon — they sell on Shopify, Walmart, TikTok. We built plumbing around that, and we’re doing the same thing now, just one layer up. We started on Amazon but we’re not going to end on Amazon. We’re building this for multichannel, for Shopify. Shopify is an interesting character because they’ve traditionally wanted to be the source of truth. They’re converging into wanting to be the ERP.

Chad Rubin:
We essentially come up against Shopify sometimes as an ERP, even though Shopify is an ERP, but they’ve added a lot of functionality to become the source of truth as well. Everybody wants to be the source of truth. Why? Because then you’re stickier, you don’t churn. Our bet is a little bit different than most companies right now. A lot of companies are like, “Here’s our integration to Claude, here’s our integration to ChatGPT.” We’re saying, “No, we don’t need an integration — we need the data ingested.”

Chad Rubin:
Everything happens behind the scenes. You don’t need to use tokens for that. We’re going to use ingestion or proper software to ingest that, and then our system becomes a layer on top with a beautiful user interface that’s purpose‑built for ecommerce. That’s the play we’re making and what we’re working toward. Shopify just launched their own integration with Claude, with an MCP where you can change the storefront. They also did the Shopify catalog — I don’t know if you saw that announcement recently.

Steve Hutt:
Yeah, I did, yeah.

Chad Rubin:
All of this is focused on betting on a horse, betting that they’re going to use Claude or ChatGPT for this. In my opinion, if you build a purpose‑built tool, it’s like using a steak knife versus a butter knife. It’s done for you already. You don’t have to be an AI engineer or a programmer or figure out how to ingest this, how to install it, or how to use GitHub. It’s like, “Hey, click a button, hire this employee.” We started with Marco. He’s the hardest. PPC is a lot harder than anything else.

Chad Rubin:
Why? Because there’s a lot of money on the line. He has to do a lot of actions, he has to analyze a lot of data.

Steve Hutt:
Yeah.

Chad Rubin:
Claudia has to arbitrate that data, observe it, understand it, understand the context of it, make the call, and coordinate it. It’s super hard. Then Oracle handles pricing. There are a lot of small jobs that need to happen. To me, the future is going to be Upwork but with agents. We’re building contractual work. You don’t have to hire these AI people full‑time. You hire them to do a specific task, and that’s part of what we’re building currently.

Steve Hutt:
It’s interesting with Oracle because the tool has to be margin‑aware as it relates to dynamic pricing. I think it’s not just finding the right price or the lowest price — it’s finding the most effective price based on the business outcomes the brand wants and what’s going on competitively in the marketplace.

Chad Rubin:
Pricing is really hard and super interesting. For example, you may want to raise price on a hero ASIN because margin is soft this quarter, but there’s an effect on your ranking position on Amazon. And when you raise price, Marco may want you to drop price because maybe your cost per click is climbing. Both of them are right inside their own world, their own AI world. I know I’m getting futuristic here, but they’re both right in a way. You have to arbitrate those differences. They’re not going to always agree.

Chad Rubin:
In fact, it might be great to have them disagree. Disagreements can create positive harmony in the process so you can make the right decision for the business. The best idea always wins.

Steve Hutt:
The other agent I see — and I love these names, that’s why I want to go through them — is Bruno. Bruno is all around demand planning and inventory, which is very interesting. It’s incredibly important to understand and forecast seasonality and sales velocity (or not) and what that means as it relates to pricing. It’s interesting how there’s a flywheel going on here. Talk a bit about Bruno and your concept around him.

Chad Rubin:
It’s the same thing. You mentioned this earlier in the call. There are all these point systems that do one thing. They do inventory, but inventory doesn’t talk to pricing; pricing doesn’t talk to PPC; PPC doesn’t talk to your CFO or QuickBooks. There’s no coordination happening.

Chad Rubin:
There’s no shared understanding of the business. It’s almost like if your right hand didn’t know what your left hand was doing.

Steve Hutt:
Yeah.

Chad Rubin:
It’s the same thing with price, but it goes much deeper. Agentic AI gave us the ability to have agents that go out and do a thing, do it with intention, and do it with a clear structure of data. They are so much smarter than us. Bruno is the next step: the supply chain. Making sure you have inventory in stock, making sure you’re not over‑buying, making sure you’re buying low, not buying high. That’s Bruno’s job.

Steve Hutt:
Yeah, it’s an important job.

Chad Rubin:
Yeah. The thing is, Bruno will never leave. He’s a teammate. He’s working 24 hours a day, reading everything. He doesn’t get tired, he doesn’t go walk his dog, he’s not checking out. You don’t have to manage him on Time Doctor or any other time‑tracking software.

Chad Rubin:
We’re essentially building guardrails. I think this is one of the most important things about AI in general. AI hallucinates, AI makes mistakes, just like humans do. The question is, how do you build guardrails around that? We’ve built some strong guardrails in the system that protect you from dumb mistakes. If you do this in ChatGPT or Claude, it’s going to make mistakes, it’s going to make dumb mistakes. It doesn’t have full context, it doesn’t connect to Fathom the way we connect to Fathom and download all the transcripts so you know exactly what’s happening beyond the data in the business. It’s checking its assumptions and checking them twice, making sure it doesn’t make a bad move for the business.

Steve Hutt:
I want to talk about data a bit because I find it interesting, and I bet you’ve noticed this with your beta testers. Maybe the data showed one thing, but then running one of these agents against that data, all of a sudden it was counterintuitive but actually the correct answer. On the surface, a human looking at the data might say, “No, this is what’s going on,” but the reality was something else was happening, and the agent’s answer was the right one to help the business and its outcomes.

Chad Rubin:
Yeah, and I think that happens. You can look at it on the pricing side. You might say, “If my competitor goes up 25 cents, then I’m going to go up by 15 cents.” These are if‑then statements, and they’re only as good as the assumptions behind them. We only have the capability to hold a specific train of thought based on our experiences and perception of the world in our brains. That’s not what happens with these agents. They look at the broad strokes — here’s all the things that are happening, here’s everything that’s happening right now: your competitor is dropping price, inventory is falling for 30 days, your conversion is dipping, your ranking position is changing.

Chad Rubin:
There’s so much data to chew on and so many decisions to make. This is what they’re really good at: ingesting and processing this data. The agent doesn’t just have one rule. It’s looking at the data and saying, “Wait a minute, here’s the sell‑through, here’s the price, here’s the ranking position.” Based on history, you raised price 20 cents and it didn’t work last time, but when you raised price here, demand snapped back, etc. It’s understanding all of this at the same time. Data expires over time. Things move so fast right now, faster than ever before, that data just expires.

Chad Rubin:
So now they don’t use fixed rules. Multiply that by every ASIN, across every channel you sell on, across every hour of the day, Steve — that’s the gap.

Steve Hutt:
Yeah. This is unbelievable. It’s amazing that we have this opportunity now, and it continues to advance as these models and your harnesses — your agents — get smarter and start working together. What I find interesting, and I want to be mindful of the different types of merchants listening today, is that you have a threshold or a fit. It seems that Profasee Ultra is built for brands starting around $50,000 a month, with at least 20 SKUs or more. Under a certain volume, you probably don’t have enough data to make all the decisions or justify the platform.

Steve Hutt:
Can you walk me through the reasoning behind that so my listeners can say, “Yeah, this is a good fit for me because I’m doing that sort of volume today”? I want to understand where the volume line is and why it’s important to be at that volume so these incremental changes the agents make are actually impactful for the business.

Chad Rubin:
Yeah. I think it comes down to “more money, more problems.” At a bigger scale, you have more challenges you need to overcome. If you’re managing one or two SKUs and making a couple thousand dollars a month, you can probably do it without an AI agent. Or maybe you can get by using some form of AI, but it’s going to be you, hands on the keyboard.

Steve Hutt:
Right.

Chad Rubin:
This is a serious platform with a lot of tech power behind it. If the velocity is too thin, it’s not going to have enough to pick up on and chew on.

Steve Hutt:
Right.

Chad Rubin:
If you’re doing over 50K, that’s the bottom threshold. If it’s climbing, we’re going to be a speedboat for you. We’re going to help you scale. The problem isn’t getting started; the problem is that you’re outgrowing your processes and the humans doing the work. I used to outgrow humans, agencies, or software tools every six months.

Steve Hutt:
Right.

Chad Rubin:
This is why we focus at least at 50K. It’s not for side hustlers. It’s for operators and brands that have shown they can sell something and now want to scale without doubling headcount.

Steve Hutt:
Right. Let’s talk about the first 30 days, because I see every new account starts in “ask me first” mode before it runs anything autonomously. I’d be frightened as an owner if it didn’t do that. Can you walk us through what the first 30 days look like, and where you find operators pushing back on this initial onboarding process?

Chad Rubin:
When you start off, you have to connect your Amazon account and connect the AI technology. It’s a click of a button — all of this is just done with clicks of a button. You have to load your cost of goods sold into the system.

Chad Rubin:
A lot of people are seeing this for the first time. The reason we built “ask me first” mode is that they want to get comfortable with the system. Anytime you use new technology — even a Tesla in self‑driving mode — you’re like, “Wait a minute, how does this work? I’ve never done this before.” And you wonder, “How do I trust that the agents won’t do something really dumb in my business?” The answer is: you don’t trust them, especially not on day one. That’s why we built AMF mode — Ask Me First mode. The agents are going to make proposals.

Chad Rubin:
They put them into what we call Mission Control — a Trello‑style card that’s awaiting your approval. Every price move Oracle wants to make, every bid change Marco wants to push, every recommendation Claudia wants to surface comes to you first. You have to give it your blessing. You approve it or decline it. When you decline it, it becomes a feedback loop back into the system; if you modify it, that also helps the agents learn your preferences. For the first 30 days — and it doesn’t have to be 30; you can switch from “ask me first” mode to “handle it” mode in 12 or 15 days — those early days are about comfort and trust. Your agents are essentially interns at this point. They’re smart as hell interns, they read every line of your data, but they’re still interns. They’re not shipping anything without you.

Steve Hutt:
Right.

Chad Rubin:
Once you start seeing the pattern, you’re like, “Oh wait, this is really good. Approve that — yeah, that’s good.” Bruno suggests a reorder quantity that’s 20% higher. You double‑check his work, and, “Oh wow, that’s right, his work is correct.”

Chad Rubin:
And that’s it. That’s the whole philosophy. You decide how much rope they get, and you decide based on how much trust they’ve earned.

Steve Hutt:
Yeah. Let’s talk about agencies, because a lot of people listening may not run brands, but they do done‑for‑you services. It’s kind of a twofold question. In a webinar you had a while ago, you made a comment: “You don’t have to fire your agency, you just need to check their work.” I think that’s one of the benefits of your platform — running Ultra in read‑only mode alongside an existing PPC agency is interesting. I’m curious what you’ve found among those who don’t necessarily want to fire their PPC agency yet, or aren’t ready to let Profasee Ultra take over.

Steve Hutt:
What are you seeing from the agency world? That’s one part. The other part is: do you see agencies adopting your platform and using it, maybe even white‑label, for their clients? I could see the power of being able to market that publicly without revealing the underlying technology.

Chad Rubin:
All right, let’s separate the questions. When we sell to brands, they’ll say, “Well, I have an agency,” and it’s super helpful for our system to get into the nitty‑gritty of their business and check the work — checking the catalog, checking PPC. It’s way smarter than the humans. We all know the agency model in general is to scale brands using the same workforce and headcount. That’s the model: they hire somebody at, let’s say, $5,000 a month, and then they scale 10 brands for that person, each paying $5,000 a month, and that’s your ROI. If we can be a wedge into these brands and show them…

Chad Rubin:
“Okay, look — you start off in ‘ask me first’ mode. You can see what’s happening in your business. You can see where the gaps are that you can exploit to make more money.” This gives us an in into the business. Once we’re in, they say, “Oh wow, this is game‑changing. I’m going to transition.” So we wedge into their business by surfacing what’s happening, and then they move forward with us, if that makes sense.

Steve Hutt:
Right.

Chad Rubin:
The second thing is agencies. At the same time we’re competing against agencies, we’re frenemies. We’re competing because we’re probably what keeps them up at night. We are the system that keeps them up at night. I like to say, “You know what, if you can’t beat us, join us.” Use our software to help you reduce headcount so you don’t have to keep scaling headcount, and you can use our system to automate the back end of the business. You can see in our footer at Profasee, we have an “Ultra for Agencies” landing page specifically built for agencies. The answer is yes, we want brands to use us; yes, brands are moving away from agencies to use us; and yes, we want agencies to use us at the same time. Now, I do want to say there’s a lot of noise. I think there’s one thing we haven’t covered: there’s a lot of noise in the space.

Chad Rubin:
Everybody and their mother is — and this is the one thing that really irks me to no end — everyone is claiming to be an AI expert, everyone’s claiming to have agents, and they don’t even know… I don’t think there’s a proper definition. You can just set up an automation and call it an automation agent.

Steve Hutt:
Yeah.

Chad Rubin:
So what are you hearing on your side?

Steve Hutt:
Yes, I am actually. I’m hearing a lot about people doing these complimentary audits — “AI visibility” — and they’re calling themselves AI experts by running a manual prompt through Claude and seeing where a brand comes up. This is on the Shopify side, where they show up in search through AI overviews or these large language models. There are a lot of third‑party connections like Ahrefs, Moz, and a few other up‑and‑coming tools like Searchable.com. People are using these tools but not talking about what the tools are. They’re just saying, “Hey, do a free visibility audit,” then getting them into a pipeline of, “Here’s what we can do to help you grow your business,” for both organic search — which bleeds over into AI overviews if you’re on the Google side — or having your data friendly enough and parsed, available through FAQs, schema, atomic answers — the things these AI engines are looking for.

Steve Hutt:
That’s what I see, at least from a content perspective.

Chad Rubin:
Yeah. I think that’s the other problem. When you go on Instagram and see photos of people doing really cool things, that’s not their true life. It’s just them showing part of their life; they’re not showing the bad things — nobody wants to see that. It’s the same thing when you go on X or LinkedIn. You open it and you’re like, “Wait a minute, everybody has AI in their bio. Everyone’s an AI strategist now. Everyone’s a growth expert.”

Chad Rubin:
They watched maybe a couple of YouTube videos to figure that out, and then they’re building courses around it. It’s just not real life. This is a big problem I have. They’re content creators — they’re not called business creators; they’re called content creators. This is what’s happening in the Amazon space. I think it’s maybe happening more on Amazon than most. Amazon has a lot of…

Steve Hutt:
Of…

Chad Rubin:
I don’t know what you want to call them — imposters or…

Steve Hutt:
I guess you can say gurus. Yeah, fake gurus.

Chad Rubin:
There’s a lot of that — a lot of charlatans who pretend to have knowledge or qualifications they just don’t have. It bothers me to no end because people will say, “How are you compared to this? This says AI‑powered on it, and you’re AI‑powered.” I’m like, “This is very different. This is not just a chatbot added to a dashboard. This is a real AI workforce — people, specifically agents, making decisions for you. They’re actually making decisions.”

Chad Rubin:
They’re reasoning, making decisions, rolling up their sleeves, and making a decision for you. They’re not just looking at a dashboard and showing you something from the data that’s already there, where you still have to do the work and decipher it. This is why I get very frustrated in the space — it’s so noisy and it’s hard to spot the fakes. I recently called somebody out in a WhatsApp chat. Somebody said, “Come check out our AI agents, but first sign up for our webinar.”

Chad Rubin:
Once you get on the webinar, then for three days you can go to this thing, and once you sign up for this course, then you can use our AI agent platform. I was like, “No, it’s just noise. It’s super dangerous and you’re pretending.” They directly messaged me and said, “Hey Chad, you’re better than this.” I said, “You — I know I maybe came across a little harsh there, but you’re better than this.” This is not a small problem. This is a credibility hit in the industry.

Steve Hutt:
I’ll talk about this briefly. I think it’s impressive that you operate the way you do. You’re running your own brand through Ultra. You’re eating your own dog food — you made a comment that you feel the product is being eaten by the company that built it. Can you talk about that? I don’t want you to give away the secret sauce of what you’re doing, but it’s interesting that you have a different lens. You’re not just a SaaS AI company. You understand the anxieties of running an ecommerce brand and what it means to be on Amazon and be effective.

Chad Rubin:
Yeah, I would say when I put out content, it hits — it hits pretty hard because I have a point of view. I’m not just creating content for clickbait.

Steve Hutt:
Right.

Chad Rubin:
Think Crucial is this brand I started 20 years ago. I was on Wall Street. My parents had a vacuum cleaner store. As it started to scale and increase, I left Wall Street — I actually got fired from Wall Street.

Steve Hutt:
Yeah.

Chad Rubin:
I started this brand in a warehouse in Harlem and it scaled significantly from there, which is how I started Skubana, how I started the Prosper Show, and how I started Profasee after I sold Skubana. I was like, “We’re struggling, our margins are getting hit. Let me do something with dynamic pricing.” I’ve always been on the cutting edge of stuff.

Steve Hutt:
Stuff.

Chad Rubin:
It’s a real business, real P&L, real inventory — there’s a lot on the line. We’re using Ultra to learn, to grow, to know what matters. I see where it’s wrong, where it’s making mistakes, what’s needed, and that helps our product get better because I’m running my own dollars through it.

Steve Hutt:
Yeah.

Chad Rubin:
I am the support ticket. I’m not waiting for a support ticket.

Steve Hutt:
Yeah, yeah. It’s amazing. When you get that frontline knowledge and feedback, plus all your beta testers, you get this feedback loop. That’s why the product is iterating, because you’re actually running it yourself. That’s very unique — not just having outside people using the platform and having a hypothesis of what you believe the platform is doing, but running it through your own significant Amazon business and retail. You’re actually on ThinkCrucial.com also. It’s pretty cool that that’s what you’re doing.

Chad Rubin:
Yeah, I think it’s a good place to be when you build from problems.

Steve Hutt:
Absolutely. What do you think some of the next steps are? I’m thinking about the people listening now who are in that sweet spot — they’re at that 50K minimum, they’ve got product‑market fit, traction, processes in place, a brand, and 20 or more SKUs. What do you believe the next steps are for those who want to think more about getting involved with Profasee Ultra?

Chad Rubin:
Well, if you’re listening to this podcast and you’re like, “Heck yeah, I love this, I need some AI agents in my life,”

Steve Hutt:
Life — yeah.

Chad Rubin:
then just email me. It’s [email protected] — P‑R‑O‑F‑A‑S‑E‑E. That’s my personal email. Let me know you heard the podcast. Let’s just say it’s June 18th, 2026, right now.

Steve Hutt:
Yeah.

Chad Rubin:
I’ll honor this for the next six months or so, but you’ll get 50% off our pricing.

Steve Hutt:
That’s amazing. I’ll make sure I put that in the show notes. That’s incredibly generous. If you’re at that level right now, you just don’t know what you don’t know. For brands listening right now, if you’re looking for a solution — it’s not just me pitching the product. I’ve written a ton of notes. I’m all in on what you’re building. It’s definitely leading edge. Even before we were recording today, you said, “Look, we’ve got a lot of things figured out. There are a lot of things that are not figured out,” but that’s part of what you’re doing with Think Crucial and your connections to all these beta testers. That’s why, even at the top of your website, there’s a link where you can apply to see if you’re a good fit. I think that’s an important distinction — it’s not for everybody.

Steve Hutt:
You’re very successful and robust, but you’re being intentional about who you’re selecting and why. Part of it is, where do you fit in Amazon’s world right now, and can you be impactful with your agents today? Then, what does the future hold with customer feedback and your own feedback from your own store?

Chad Rubin:
Totally. From 2015 to 2025, there was this software playbook — 10 years. It was a really nice ride.

Steve Hutt:
Yeah.

Chad Rubin:
But those days are over now. There is no playbook for what we’re doing. Like you said, a lot of Profasee Ultra is not figured out. I just shared with you where we’re going and how it works.

Chad Rubin:
We’re still working out stuff: how much autonomy do you give Claudia? What happens when agents disagree? There’s a lot — there’s a whole list. The honest part is that it still needs to be taken seriously. People who say, “Oh, AI is not going to change the world, it’s overrated,” — the answer is, it’s actually not. This is very early. This is the dumbest the models will ever be. The models are still early. If you think about it, even with Fable and the whole story with Fable — which is hopefully going to be rectified soon — Fable 5 with Mythos, that model is supposed to be a game‑changer, but that’s just… we’re still in the first inning of changing the game.

Steve Hutt:
Right. You see it with Elon too — his IPO and his recent purchase of Cursor for $60 billion has proven there’s vibe coding and coding in general. We thought Claude Code owned the market, but holy smokes, if you have a $60 billion acquisition, clearly we know where the market is and where it’s continuing to iterate.

Chad Rubin:
Wild times that we’re living in.

Steve Hutt:
Yep, this is great. Once again, Chad, thank you for recording today. Like I said, I’ve written a lot of notes. I’m going to make sure we get this out into the wild and share it. Once again, if you’re listening and you run an Amazon brand and you’re at this 50K with 20 SKUs or more, email Chad directly — [email protected] — or go to Profasee.com and have a look. There’s an “Apply” button in the corner. Mention the show, and I appreciate the special offer for 50% off for six months so people can actually see what the platform can do for them.

Steve Hutt:
I guess we’ll see you in another six months or so and talk about what new agent you’ve hired, for sure.

Chad Rubin:
Thank you so much. I’m super grateful to be here with you.

Steve Hutt:
Yeah, this is lovely. All right, have yourself a great afternoon. Well, that’s it for today’s episode. I’d like to thank you personally for being a loyal listener of eCommerce Fastlane. It’s my hope that this podcast is offering you a ton of value through growth strategies, tactics, and exclusive insider tips on the best Shopify apps and marketing platforms, all with my personal goal to help you build, manage, grow, and scale a successful and thriving company powered by Shopify. Thanks for investing some time today and listening to the show. I’m so proud and excited that you have a growth mindset and are a constant learner. I truly appreciate you and your entrepreneurial journey.

Steve Hutt:
Enjoy the rest of the week and keep thriving with Shopify.