🇺🇦 Evgeny Shadchnev

evgeny.coach

I coach founder CEOs and their boards through leadership transitions: stepping down as a CEO or stepping into the CEO role (even if the founder had the title for years).

Founder Coach

Makers

Co-founder and Board member

Makers Academy is an intensive, full-time, 12 week course in web development. We help our students to change careers and become junior software developers and apprentices. We help companies hire great junior programmers.

All Together

Powered by a collective of CEOs from some of the UK's most successful companies, All Together is a pro bono advisory platform offering 1-2-1 advice, peer learning events, and impactful online content to help you build a better business.

Volunteer Advisor

Ada Ventures

Ada Ventures finds and funds extraordinary talent-building breakthrough ideas for the hardest problems we face. We invest across climate, economic empowerment and healthcare & ageing.

Venture Scout

Femme Palette

Mentor

Forward Labs

Forward Labs is a startup lab. Using lean startup methodologies, we validate ideas, rapidly build and test products and hire teams around those products that have solid traction.

Founding partner of Forward Labs, responsible for overall strategy and management, ideation, development of MVPs (Ruby, Rails, iOS, etc), hiring developers and entrepreneurs etc. During the first year we grew the team from 3 to 18 people and worked on Dropwines.com, MakersAcademy.com, PocketShop.com, AppearHere.com and several other projects. In 2014 Forward Labs was merged into a newly created investment fund Forward Partners.

Founding Partner

Forward Internet Group

Technical co-founder of InvisibleHand (http://www.getinvisiblehand.com/), the most popular price comparison browser extension in the world, which since its launch found over 1 billion dollars in savings on products, flights and car hire for hundreds of thousands users.
I led the development of the project for the first two and a half years. The backend written in Ruby and JS maintained an consistent, up-to-date database of dozens of millions of items across hundreds of retailers serving real-time prices to the users.
Ruby / Rails / Sinatra / JS / advanced regular expressions / AWS / MySQL

The company was sold to Skimlinks in August 2013.

Tech co-founder at InvisibleHand

Kappa Prime

Kappa Prime is a startup aiming to help research community to keep abreast of latest scientific developments.

Ruby / Rails / HTML / JS / CSS.

Co-founder

Etera UK Ltd.

Etera UK is a part of international group of companies specializing in web development. We provide a complete range of web solutions: market and competitors analysis, web design, web programming, UI expertise, testing, internet marketing and other services. We develop corporate web-sites, product web-sites, personal web-sites, intranet solutions, billing systems, issue tracking systems etc. In most projects we use our own purpose-built CMS - Sapphire.

Managing Director

Kodofon

Developed a part of a bespoke IEEE 802.16 (WiMax) simulator. Researched the use of error-correction codes in the wireless communications.

C++ / Matlab.

Wireless engineer

No Founder CEO should feel trapped in their job | Author of Startup CEO Succession | ex-CEO at Makers

Feeling trapped as a founder CEO, uncertain about whether to stay or step down?

I’ve been in your shoes, having built and led a successful startup, Makers. I’ve navigated the challenging decision to step down and wrote “Startup CEO Succession: A Founder’s Guide to Leadership Transition” to help others like you.

I coach founder CEOs and their boards through this critical process. Whether you choose to stay and enhance your leadership skills or decide it’s time to pass the torch, I offer personalized coaching every step of the way.

𝐒𝐭𝐚𝐲: I help you become the best startup CEO you can be.
𝐋𝐞𝐚𝐯𝐞: I guide you through a smooth transition, selecting your successor and your next steps.

DM me to turn your uncertainty into clarity and confidence. Let’s unlock your full potential, no matter the path you choose.

I coach founders who'd like to be the CEO their startup needs right now. Together, we navigate the uncertainty, grow in self-awareness and meet the challenges of startup growth.

As a software developer 20 years ago, I internalised the wisdom that waterfall is bad and agile is good. Planning complex things in advance doesn't work; being responsive and iterating quickly works much better.

As I'm learning the world of AI-assisted development, I find myself going back to waterfall principles: planning complex documents, splitting them into development phases, and then implementing them. This intuitively feels more productive than being agile with AI. But why? The same approach clearly doesn't work with people.

I think, on reflection, there are two differences:

It takes far more time to plan anything complex with people. Therefore, if I try to produce a complex document that tries to describe too much complexity taking input from multiple stakeholders, I will fall flat on my face. So being agile and responsive wins.

With AI, if I'm the product person, the engineer manager, the coder, the tester, and the CEO all in one, I can do dozens of iterations planning a complex feature expressing every detail in long and detailed documents in just a couple of hours. This would have taken weeks with real human teams.

And then when it comes to implementation, since the speed is so much faster than human development, I can go through the implementation phases within a day on something fairly complex which again would have taken weeks in the pre-AI world.

So my actual process looks like waterfall at the speed of Agile, which sounds like a contradiction because it's impossible in the human world but is possible in the AI world. And the reason I'm investing so much time in the specification is because if the documentation and the spec is good, then AI is going to do a decent job implementing it. And therefore I am reducing the likelihood of problems introduced by AI coding by investing heavily into thinking everything through up front in dialogue with AI.  

I'm not saying I'm discovering any truths about AI-assisted development. We're all figuring it out as we go. I'm just describing what I'm learning in the moment.

Evgeny Shadchnev

One workflow I found very powerful with AI is asking AI to interview me to improve the quality of the documents I'm writing.

For example, I might create the document (probably with the help of AI). Then I give it to AI again (or maybe even to a different system or two) and ask them to keep asking me questions to clarify what the document is trying to say until it's perfectly clear.

Even after one chatbot says that everything is clear, you can give it again and again to the same chatbot or a different one, and it'll find new clarifying questions to ask. Stop not when it tells you its clear but when you decide it's clear.

Since I can answer the questions by voice, it becomes much easier than typing. The process of clarifying my thinking in a dialogue leads to a much higher quality result with far less effort compared to writing the entire document by hand as I might have done a year ago.

If you haven't tried it yet, give it a go.

I've been preparing for a talk at Full Stack Founder AI Bootcamp, and I keep noticing patterns in how even brilliant people misunderstand AI. I've been guilty of these myself.

𝗪𝗲 𝘁𝗵𝗶𝗻𝗸 𝗔𝗜 = 𝗰𝗵𝗮𝘁𝗯𝗼𝘁𝘀
Most AI discussions never leave the chat interface. But that's like equating the internet with email. While we're typing prompts into ChatGPT, AI is discovering antibiotics, folding laundry, and creating virtual worlds. The comfortable chatbot frame might be making us miss the forest for one particularly charismatic tree.

𝗪𝗲 𝗱𝗶𝘀𝗺𝗶𝘀𝘀 𝗶𝘁 𝗮𝘀 "𝗷𝘂𝘀𝘁 𝗽𝗮𝘁𝘁𝗲𝗿𝗻 𝗺𝗮𝘁𝗰𝗵𝗶𝗻𝗴"
Sure, philosophically we can debate whether pattern matching constitutes "real" understanding. But when AI writes and executes code that runs actual systems, or diagnoses medical conditions that save lives, the distinction becomes academic. What matters isn't whether it truly understands—it's whether it works. And increasingly, it does.

𝗪𝗲 𝗷𝘂𝗱𝗴𝗲 𝘁𝗼𝗺𝗼𝗿𝗿𝗼𝘄 𝗯𝘆 𝘁𝗼𝗱𝗮𝘆'𝘀 𝗹𝗶𝗺𝗶𝘁𝗮𝘁𝗶𝗼𝗻𝘀
Three years ago, ChatGPT could barely write paragraphs. Today it's a better programmer than most humans. The length of tasks AI can complete autonomously is doubling every seven months. By 2027—closer than ChatGPT's launch—we might have AI handling full workdays. Yet we keep thinking linearly about an exponential technology.

𝗪𝗲 𝗮𝘀𝘀𝘂𝗺𝗲 𝗼𝘂𝗿 𝗲𝘅𝗽𝗲𝗿𝗶𝗲𝗻𝗰𝗲 𝗶𝘀 𝘂𝗻𝗶𝘃𝗲𝗿𝘀𝗮𝗹
The coder thinks AI is about generating functions. The writer sees a content machine. The therapist sees an empathy engine. We're all touching different parts of the elephant, convinced our piece represents the whole. No one has the complete picture.

𝗛𝗲𝗿𝗲'𝘀 𝘄𝗵𝗮𝘁 𝗵𝗮𝘂𝗻𝘁𝘀 𝗺𝗲: Cognitive scientist John Vervaeke points out that consciousness performs "relevance realisation"—filtering infinite possible attention into what matters. AI inherits our sense of relevance through training data. But what happens when AI trains on AI-generated content? Without human input about what matters, might it lose its grounding in reality?

I don't know. I don't have neat conclusions. The more I work with AI, the less certain I become about what it is and where it's going. Maybe that's the appropriate response: not certainty, but curiosity.

𝘐'𝘮 𝘴𝘱𝘦𝘢𝘬𝘪𝘯𝘨 𝘢𝘵 𝘍𝘶𝘭𝘭 𝘚𝘵𝘢𝘤𝘬 𝘍𝘰𝘶𝘯𝘥𝘦𝘳 𝘈𝘐 𝘉𝘰𝘰𝘵𝘤𝘢𝘮𝘱, 𝘓𝘰𝘯𝘥𝘰𝘯, 𝘚𝘦𝘱𝘵 29-30. 𝘜𝘴𝘦 𝘤𝘰𝘥𝘦 𝘌𝘝𝘎𝘌𝘕𝘠100 𝘧𝘰𝘳 𝘥𝘪𝘴𝘤𝘰𝘶𝘯𝘵.

I can't decide which of these is more crazy.

On balance, it's the one of the left. However we get out of the mess the world is in, I’m sure trillionaires aren’t part of it.

My psychoanalyst once told me I never get stuck in my dreams; I always find a way out. I took it as a compliment. How resourceful!

Then I started getting stuck. Waking up with my heart racing. Entering abandoned buildings floating in voids, filled with trapped souls and dead children. It felt like failure. It was actually progress.

What seemed like clever problem-solving was really just avoiding the parts of myself I couldn't face. Getting stuck meant I was finally approaching them.
This insight transformed how I lead.

As a CEO, I used to dread the emotionally loaded conversations. Layoffs during COVID. Looking someone in the eye, knowing their spouse is pregnant, telling them their job doesn't exist.

I kept searching for the pain-free approach. The perfect playbook.

But some situations aren't meant to be easy. The leaders who dissociate from this pain, who pursue corporate goals at the cost of human wholeness, become the worst kind of leaders.

The adult question isn't "How do I avoid pain?"

The adult question is "How do I stay present with it?"

When your team sees you choosing what's hard and important, bringing your full awareness there, standing still even with your hair on fire, they understand something their minds can't articulate. If you can do it, maybe they can too.

Progress often feels like regression. The racing heart, the moment of being stuck, the painful conversation—these aren't obstacles to growth.

They're signs you're finally approaching what matters most.

Getting stuck, it turns out, is how we get free.

(Link to the full essay in the comments.)

> AI agent, look at today's commits and estimate how many days this would have taken one experienced developer if they were doing this work by hand without AI assistance?

Grok: 1-2 weeks of solid development work

GPT5: 6–7 developer-days

o3: roughly two work-weeks

Claude4: 8-10 full development days

That's 5-10x (I ran it on my own codebase just now).

If your business is planning to outperform the competition who are using AI-assisted development without doing it yourself, you must believe one of two things:

- what you see above can't be true
- your competitors aren't smart enough to produce good code using AI

I really wouldn't bet on either.

Taplio

Evgeny Shadchnev PCC's Linkedin Analytics

What is Evgeny talking about?

's Best Posts (last 30 days)

Want to drive more opportunities from LinkedIn?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Evgeny Shadchnev PCC's Linkedin Analytics

What is Evgeny talking about?

's Best Posts .css-qqfgvy{font-size:var(--chakra-fontSizes-md);}(last 30 days)

Want to drive more opportunities from LinkedIn?

Evgeny Shadchnev PCC's Linkedin Analytics

's Best Posts (last 30 days)