About me

School years and adolescence

I started my training at an art school, where I diligently went after regular classes for 6 years, planning to become an architect. Once I finished art school, I never drew again in my whole life.

At the same time, I turned out to be good at Physics and Math, so I started competing in Olympiads, and even won at the national level, but didn’t go for the international because of the politics between prestigious high schools.

I worked every summer starting from 13, mostly doing bullshit like posting ads on buildings or distributing flyers; sometimes, I got to work on construction sites, and I loved it because it paid the best. My dearest memory is the manufacturing of concrete blocks, where I worked together with my friend, just two 15-year-olds under the baking summer sun, all in concrete, sand, and water, forming 25kg blocks with a tamping machine that would occasionally electrocute us. We had to commute 2 hours outside of the city, trying to get there at 6-7 am before the heat sets, working full day until we collapsed powerless onto a pile of sand. I worked there for a month until my partner stopped showing up, and I was left alone with our bricks and sun. I did one more month on my own, and after that, I never considered any work in a comfortable, climate-controlled office as specifically “hard”.

I got the highest score on the national graduation exam in Physics (sort of SAT) and got accepted to top-1 university in Ukraine in the Physics department. I didn’t know anything about the world, and school years were fun for me, so the plan was to become a teacher.

University years

It almost happened when I started working independently as a Physics and Math tutor, although it was only in my 3rd year, because before that I didn’t have time for anything other than studies.

At the same time, I got an entrepreneurial itch, and after a failed attempt to launch a romantic mobile app and a humiliating attempt to sell app development to a water delivery company, I started my first “business”.

I still remember the moment I got the idea, I couldn’t sleep, the shit was genius—I would resell plush alpacas from China. I created not even a store, but a social media page that I would grow purely organically, would stock alpacas from Aliexpress in my dorm, and would resell them with a 50% margin. That’s where I learned that getting rich from a business is a scam—it was just a lot of work for very little money, but given that I didn’t spend anything, it ended up a profitable venture.

I also got interested in modeling and even did a couple of gigs as a model. It was mostly driven by curiosity, cool IG pics, and ego.

Research

Physics

In the short breaks between plush alpacas and catwalks, I did cancer research. Since I pursued an Experimental Physics specialization, it was easy to get involved in faculty research, mostly to gain some cool experience for my future CV, and boy, it delivered. I got to play with liquid nitrogen, lasers, and spectrometers. My research was focused on silver nanoparticles paired with specific organic molecules for cancer treatment. A lot of late evenings in the lab, a lot of photonics, and atomic physics. But it was fun! And easier than concrete blocks, 100%.

Experimental Psychology and AI

I did my Master's in France, Spain, and Norway. Computer Science was kinda easy and exciting after Physics, so I quickly started going the extra mile and ended up solving cross-disciplinary challenges that I formatted into research papers. I started going deeper into Machine Learning and Deep Learning, and because of the cross-disciplinary nature of my program, I started applying it to Vision and Image Perception, which allowed me to work for free in Experimental Psychology labs in KU Leuven (Belgium) and the University of Oxford (UK).

Being abroad and doing something beyond graphs was very exciting, and I was hungry for any real-world experience. I ended up involving myself in multiple projects in different labs: did more Physics research in France, did EEG and eye-tracking in the UK, and trained Generative AI models in Norway. A lot of papers I wrote as a solo author in my free time, for some, I used school resources, and asked feedback from my professors, but still mostly worked alone, constantly training models on my laptop or school servers. By the end of year 2, I had 14 papers in total, and presentations at conferences like CVPR, which paved my path to the next destination—Facebook AI Research.

Facebook AI Research

At Meta, I worked in the Multimodal team, tackling challenges at the intersection of Vision and Language (Image Captioning, Visual Question Answering, Reading Comprehension). We collected and annotated the first public dataset of images with text in it, and proposed a transformer-based SOTA model for visual question answering with reading comprehension (it was later added to ChatGPT in 2024). We received a prestigious oral presentation at ECCV, and it’s my most cited paper to date.

Research operates in a very delayed gratification manner: the biggest proof of your work’s value is others using it and building on top of it, which manifests in citations. Now, in 2025, my papers have collected 1000+ citations from people from Harvard, MIT, Google Research, and Microsoft, but not like it even matter anymore.

I knew I didn’t plan to stay in Research and pursue a PhD, so I cut my teeth on product analytics for Facebook products, and got offers from Google and Apple, but eventually couldn’t get an H-1B visa, so I realized it would be a great opportunity to try myself in startups. This year, I celebrated my fifth year of “trying”.

Startups

During my exploration stage, I, with the help of good friends, developed an app to create deepfakes of talking heads, worked on a pronunciation correction app, started an AI news media page and wrote an interactive Deep Learning course, built a no-code platform for finding English language practice partners, predicted CTR of YouTube thumbnails, tried to turn my research in image memorability into a product, and, most exciting of all, explored decoding of EEG brain signals into images. I later sold the above-mentioned media page that now has 80k+ followers.

Suggestr

I ended up in Singapore in an incubator program where I met my co-founder. After a few more iterations, we landed on using the power of Multimodal AI and collaborative filtering in ecommerce, to predict what a customer is likely to buy next and provide “Amazon-quality recommendations to mom-and-pop stores”.

We’ve built an app, got featured on Shopify’s app store homepage, and got it adopted by 100+ brands for whom we generated more than half a million in additional revenue, which got us into Y Combinator, the most prestigious startup program in the world. But after the batch, we realized that the market ceiling is actually too low to fulfill our venture ambitions, and within a week, we returned the money to the investors, to start clean, on a bigger market.

Slise

Without taking any breaks, driven by the hunger to make it right, I started a new company called Slise. Initially intended as a blockchain analytics company, we decided to build an application layer for our data and ended up building “Google Ads for the crypto market”.

Starting from scratch, we built hyper-personalized targeting algorithms based on onchain data, the ad tech needed to serve ads, and an ads manager to track and test campaigns. We’ve grown our network to 700+ publishers and 7,000,000 users seeing our ads, and worked with such big names as PayPal, Revolut, and Metamask on the buy side.

Despite the global startup crisis and the crypto market constantly going down since the day we started, we’ve reached profitability, and after two long years, decided to sell the business to a larger media company. It took us 6 months to negotiate and formalize the deal, and another 4 months to transition the tech and hire and train the new sales team.

Dise

I’m not good at sitting still, so while working on the deal and still running Slise, we started building a new product, this time driven by our own pain and insights about the market. Our experience in crypto ad sales opened our eyes to how outdated traditional sales stack is when it comes to instant messaging.

So, we’ve built Dise—an AI-native CRM for the post-email era. We combined CRM, task management, sales automation, and outreach in a single lightweight interface tightly integrated with messaging apps. It was used by 500+ teams, including Adidas, TON Foundation, Polygon, a16z Crypto, and MEXC, and so far was the most beautiful and sticky product I’ve designed.

But our venture ambitions didn’t allow us to relax—in 2025, when AI was changing the way we live and work, we couldn’t let our AI expertise collect dust, so we made a hard decision to sell the product and pursue bigger challenges. Once again, we bent reality to force an acquisition and got a positive outcome from this not-so-long story.

Final thoughts

I write it as a 29.5-year-old confused about his place in the world, in part to figure out for myself if I have any expertise to claim in my name. I’m not particularly good at anything specific, and my corporate career prospects are murky. I don’t tolerate bullshit and fake people. I haven’t stopped since I started working, and I’m chronically tired. All 3 businesses I brought to the world continue to grow and glow beyond me.

As a message to my future self, quoting Jimmy Eat World:

It just takes some time
Little girl, you’re in the middle of the ride
Everything, everything’ll be just fine
Everything, everything’ll be alright, alright

(bonus) Investing

I feel it’s worth mentioning here, because so far, in absolute numbers, it’s made me more money than any specific venture mentioned above.

I’ve never had a habit of spending. In my school years, my free cash flow was what I could save from my bus allowance by walking instead. In the Uni, I lived for a scholarship of $40/mo, and felt like the richest man in the world. So, as soon as I started making any money, I started saving it, and once the amount became too stupid to just be lying around—started investing everything.

More than anything, I was eager to learn while I had not much to lose, so I invested across a range of assets:

—In the public market, I made a couple of good picks from the bottom—I made 7x on $META, 5x on $SPOT, 10x on $HOOD, and 12x on $NVDA. I was always cautious of not doing anything stupid, and so far, only 4 out of my 28 positions are in the red, and the whole portfolio is up more than 2x.

—In the private sector, my network in startups allowed me to invest in cool projects of my friends, so I made 8 angel investments across AI, physics, fintech, crypto, dating, and ecommerce. I mostly expect to lose money there, but so far, two of them were consecutively backed by a16z, and two got into YC, so I want to believe I got a taste in startups.

—The most interesting for me was late-stage investments in startups through secondaries—I made 7 investments, and already had two exits: Weights & Biases (MLOps) was acquired by CoreWeave at 3x, while Circle (stablecoins) made 10x at IPO—so far so good.

—I’m too disillusioned about crypto to gamble with shitcoins, but being in the market long enough, I had a chance to buy $BTC at $18k and $SOL at $9 to hold it and play a bit on volatility, instead of being exit liquidity for yet another token.

I don’t expect ever to do investing professionally, but at the same time, I see how, over time, it becomes a more and more substantial part of income, and in the end game, will be the sole basis for my retirement, so I want to keep growing my expertise there.