
Sign up to save your podcasts
Or


He runs five businesses with zero employees, and he's never written a line of code in his life.
SummaryDmytro "Dima" Negodiuk calls himself a fractional AI officer, a title his own AI agent came up with. From New York, he runs five-plus businesses across ecommerce, B2B distribution, retail, education and AI consulting, all orchestrated by a single AI agent, and without a marketing, IT or sales department behind him. In this conversation Matt and Dima get practical about where an ecommerce founder should actually begin with AI, why Claude Code is the on-ramp even for people who've never touched the tech, and how to point it at the biggest problem in your business rather than treating it like a toy.
They dig into the difference between playing with AI and deploying it, the 24/7 use cases that hand small stores an unfair advantage, and why a human still has to stay in the loop. Dima shares his AI call centre running on its 44th version, the meeting "dossier" that knows what golf game you lost twelve years ago, and the overnight autonomous work that means he checks his AI before he brushes his teeth. Matt offers his own mobile-optimisation story, where one podcast transcript turned into a scoring tool and lifted mobile conversion by over 400%.
(01:00)
Dima describes a model that would have sounded impossible a few years ago. He runs more than five businesses across ecommerce, B2B distribution, retail, education and AI consulting, all sitting on top of one AI agent. As an entrepreneur who has never coded, he's replaced the departments he used to run with AI.
"I never wrote a line of code in my life. Never. So I had like marketing system, marketing department. I have IT department, sales department, everything. Now I'm like sole entrepreneur." — DimaHe points out that the title isn't just his invention. The big AI labs now hire for the same role under a different name, the forward deployed engineer, at starting salaries north of $300,000. The word that matters in both is "deploy".
"Main word deploy. So you need not only to know how to, for example, create it, you need to know how to deploy it and all the system need to work." — DimaWhy Claude Code Is the Place to Start(22:00)
For founders who feel AI is passing them by, Dima's advice is blunt and encouraging at once. Start today, start with Claude Code, and don't be embarrassed about how little you know.
"When I saw my first video and it says open terminal on your MacBook, I ask ChatGPT like web version, what is terminal? And I'm not shy about it." — DimaA few practical pointers from the conversation:
Matt agrees it's a genuine on-ramp, and adds that the choice between tools is a bit like Apple versus Android. Dima uses Claude Code for ecommerce; friends building crypto trading bots reach for systems that fire requests many times a second instead. Stick to what fits your niche.
Start With the Problem You Actually Have(27:00)
The most useful starting point, both agree, is the biggest problem in front of you. No product? Ask for product ideas. A product that won't sell? Ask for marketing ideas. Then check the output, because AI gets things wrong.
Matt tells the story that makes the case. After Adam Pearce came on the show to talk about mobile optimisation, Matt fed the episode transcript to Claude, had it extract the key ideas, research and score each one, and eventually build a tool that audits a website against current best practice.
"Mobile conversion on our website so far is up almost 500%. It's definitely over 400% increase, um, on mobile conversion. You're talking about hundreds, if not millions of pounds worth of difference." — MattDima frames the same idea around two questions worth asking of any business. Where are the problems AI could help with, and what are the repeated tasks you could automate? A one-off phone call is quicker to make yourself. A task you do every day is worth handing over.
"Now your knowledge and your ideas it's your best benefit because everything else will Claude do for you. But one thing he can't do, it's like, I think, to dream and to generate really brilliant ideas." — DimaThe 24/7 Use Cases That Give Small Stores an Edge(31:00)
This is where the conversation gets concrete. Dima runs an AI call centre built on 11Labs, now on its 44th iteration, with Claude analysing each call and suggesting improvements. For his granite pavers business, where staff sometimes step away to help with a forklift, it means no missed call ever becomes a lost customer.
"If you will not answer immediately, the client will go to someone else who will answer immediately." — DimaThe point isn't novelty, it's the round-the-clock cover that small companies usually can't offer. A customer with toothache at 2am who can't reach a dental practice will phone the next one. The business that answers wins.
He also describes a meeting-prep skill he calls the "dossier", which spends hours researching a person from open sources and then advises on how to approach them. And for mid-sized businesses drowning in mismatched spreadsheets and systems, he makes the case that you don't need to rip anything out. You put AI on top, let it find the mess, and let it clean the data.
"It's how it works that each day you have some kind of ideas, new ideas, and you wanna make it live... It's like I call it like some kind of AI addiction." — DimaAI Is a Co-Pilot, Not a Pilot(14:00)
For all the enthusiasm, both are clear that a human stays in charge. AI hallucinates, apologises, and moves on, but it doesn't carry the consequences. Matt mentions a man fined £30,000 by the tax office after letting AI file his taxes incorrectly. The AI says sorry. The person pays the bill.
"This is why I often talk about AI needs to be a copilot. It can't be in charge." — MattDima ends on a characteristically direct note. The pace of change is so fast that even the labs won't forecast more than a couple of months ahead, so the only sensible move is to get in the water and start.
"You just need to sit and you just need to understand that if you will not spend at least 10 hours per day learning AI now, AI will— someone will control you who is now doing this." — DimaToday's GuestToday's GuestToday's guest: Dmytro "Dima" Negodiuk
Company: Negodiuk
Website: https://negodiuk.ai
LinkedIn: Connect with Dima on LinkedIn
Episode link: https://www.ecommerce-podcast.com/hes-never-written-a-line-of-code-in-his-life
By Matt Edmundson5
1010 ratings
He runs five businesses with zero employees, and he's never written a line of code in his life.
SummaryDmytro "Dima" Negodiuk calls himself a fractional AI officer, a title his own AI agent came up with. From New York, he runs five-plus businesses across ecommerce, B2B distribution, retail, education and AI consulting, all orchestrated by a single AI agent, and without a marketing, IT or sales department behind him. In this conversation Matt and Dima get practical about where an ecommerce founder should actually begin with AI, why Claude Code is the on-ramp even for people who've never touched the tech, and how to point it at the biggest problem in your business rather than treating it like a toy.
They dig into the difference between playing with AI and deploying it, the 24/7 use cases that hand small stores an unfair advantage, and why a human still has to stay in the loop. Dima shares his AI call centre running on its 44th version, the meeting "dossier" that knows what golf game you lost twelve years ago, and the overnight autonomous work that means he checks his AI before he brushes his teeth. Matt offers his own mobile-optimisation story, where one podcast transcript turned into a scoring tool and lifted mobile conversion by over 400%.
(01:00)
Dima describes a model that would have sounded impossible a few years ago. He runs more than five businesses across ecommerce, B2B distribution, retail, education and AI consulting, all sitting on top of one AI agent. As an entrepreneur who has never coded, he's replaced the departments he used to run with AI.
"I never wrote a line of code in my life. Never. So I had like marketing system, marketing department. I have IT department, sales department, everything. Now I'm like sole entrepreneur." — DimaHe points out that the title isn't just his invention. The big AI labs now hire for the same role under a different name, the forward deployed engineer, at starting salaries north of $300,000. The word that matters in both is "deploy".
"Main word deploy. So you need not only to know how to, for example, create it, you need to know how to deploy it and all the system need to work." — DimaWhy Claude Code Is the Place to Start(22:00)
For founders who feel AI is passing them by, Dima's advice is blunt and encouraging at once. Start today, start with Claude Code, and don't be embarrassed about how little you know.
"When I saw my first video and it says open terminal on your MacBook, I ask ChatGPT like web version, what is terminal? And I'm not shy about it." — DimaA few practical pointers from the conversation:
Matt agrees it's a genuine on-ramp, and adds that the choice between tools is a bit like Apple versus Android. Dima uses Claude Code for ecommerce; friends building crypto trading bots reach for systems that fire requests many times a second instead. Stick to what fits your niche.
Start With the Problem You Actually Have(27:00)
The most useful starting point, both agree, is the biggest problem in front of you. No product? Ask for product ideas. A product that won't sell? Ask for marketing ideas. Then check the output, because AI gets things wrong.
Matt tells the story that makes the case. After Adam Pearce came on the show to talk about mobile optimisation, Matt fed the episode transcript to Claude, had it extract the key ideas, research and score each one, and eventually build a tool that audits a website against current best practice.
"Mobile conversion on our website so far is up almost 500%. It's definitely over 400% increase, um, on mobile conversion. You're talking about hundreds, if not millions of pounds worth of difference." — MattDima frames the same idea around two questions worth asking of any business. Where are the problems AI could help with, and what are the repeated tasks you could automate? A one-off phone call is quicker to make yourself. A task you do every day is worth handing over.
"Now your knowledge and your ideas it's your best benefit because everything else will Claude do for you. But one thing he can't do, it's like, I think, to dream and to generate really brilliant ideas." — DimaThe 24/7 Use Cases That Give Small Stores an Edge(31:00)
This is where the conversation gets concrete. Dima runs an AI call centre built on 11Labs, now on its 44th iteration, with Claude analysing each call and suggesting improvements. For his granite pavers business, where staff sometimes step away to help with a forklift, it means no missed call ever becomes a lost customer.
"If you will not answer immediately, the client will go to someone else who will answer immediately." — DimaThe point isn't novelty, it's the round-the-clock cover that small companies usually can't offer. A customer with toothache at 2am who can't reach a dental practice will phone the next one. The business that answers wins.
He also describes a meeting-prep skill he calls the "dossier", which spends hours researching a person from open sources and then advises on how to approach them. And for mid-sized businesses drowning in mismatched spreadsheets and systems, he makes the case that you don't need to rip anything out. You put AI on top, let it find the mess, and let it clean the data.
"It's how it works that each day you have some kind of ideas, new ideas, and you wanna make it live... It's like I call it like some kind of AI addiction." — DimaAI Is a Co-Pilot, Not a Pilot(14:00)
For all the enthusiasm, both are clear that a human stays in charge. AI hallucinates, apologises, and moves on, but it doesn't carry the consequences. Matt mentions a man fined £30,000 by the tax office after letting AI file his taxes incorrectly. The AI says sorry. The person pays the bill.
"This is why I often talk about AI needs to be a copilot. It can't be in charge." — MattDima ends on a characteristically direct note. The pace of change is so fast that even the labs won't forecast more than a couple of months ahead, so the only sensible move is to get in the water and start.
"You just need to sit and you just need to understand that if you will not spend at least 10 hours per day learning AI now, AI will— someone will control you who is now doing this." — DimaToday's GuestToday's GuestToday's guest: Dmytro "Dima" Negodiuk
Company: Negodiuk
Website: https://negodiuk.ai
LinkedIn: Connect with Dima on LinkedIn
Episode link: https://www.ecommerce-podcast.com/hes-never-written-a-line-of-code-in-his-life

8,748 Listeners

216 Listeners

889 Listeners

736 Listeners