I build with AI. I coach founders.

I have spent twenty years making internet products. Programmer. CTO. Head of product. Repeat founder. In November 2025 I started building with AI. I am newer at AI than most people writing about it. I am more experienced at building than most people doing it.

Coaching

I work one-on-one with founders, solopreneurs, and product leaders who want a thinking partner who has built before. I am ICF-certified (ACC) with 150+ hours of practice and twenty years as an operator. The combination is rare. Founders coaching →

I also keep two cup-of-coffee coaching slots open each week for general clients, plus standard sessions for anyone working through a personal question. General coaching →

Products and projects

Life OS is the personal operating system I am building with AI. It runs my dating, health, journaling, projects, and coaching practice through one assistant. Markdown as database, Git as audit trail, LLM as the brain. It is the tool I use most days, and the tool that produced this site.

1clarity.app is a small tool I built solo with AI in November 2025 to help users determine their core values. The first product I shipped without writing a line of code by hand. From 1clarity I learned that AI building is a production-ready way to deliver software. I had been skeptical about that. I no longer am.

alexterekhov.com is the personal website you are reading. I built it on top of my Life OS, using Claude Code as the implementation tool. My second AI-built deliverable, also shipped without writing a line of code by hand.

AvtoTochki is the first map-based directory of car services in Russia and a marketplace connecting service businesses with consumers. I founded it after a service station scammed me. I was the author of the idea, the initiator, and the main moving force. I convinced two other founders to leave their jobs and commit years of their lives to it with me. We grew it to 1.2 million monthly active users. The product was an early at-scale application of programmatic SEO to local services. From AvtoTochki I learned how to start a business from scratch and how to sell it; it was the first real exit of my life. I learned how to build self-sustainable systems with positive feedback loops that run for years without maintenance. The project lived in zombie mode for a decade with almost no outages, and grew the whole time. I learned how to coordinate a small fully-remote team and the basics of Scrum and agile. I also learned how to handle founder conflicts, and the importance of options and cliffs in equity agreements.

Hipdir is a universal map-based directory and marketplace across every category we could parse, which I co-founded. Programmatic SEO across geography, service, and brand generated tens of millions of pages and 1.4 million monthly active users. The business ran at 90% margins. From Hipdir I learned how fast the competitive landscape on the internet can shift. What was possible to build a few years earlier was not possible by the time we hit our ceiling. The amount of competition in the market you are entering matters as much as the idea.

api.mail.ru was the technical API platform for game and application developers I created from scratch as head of product at Mail.ru Group. I led the product, the development, and the market creation. Game developers existed but they were building standalone games. My job was to convince them to build inside our social network, which was the third-largest in Russia by users at the time. We grew it to second place by users and second by revenue, and the platform reached tens of millions in annual revenue. I learned how to create a developer market from zero, how to drive an organisational shift toward agile and cross-functional teams inside a thousand-person company that was built for waterfall, and how to manage a growing team.

Futubra was a cross-platform social network I launched in seven months as head of product at Mail.ru Group. I hired the entire team from scratch and shipped on web, iOS, and Android. It was one of the first mobile-first products in Russia and the first one inside the company. The product worked. The market was already taken. From Futubra I learned how to ship at speed inside a major company, how to publicly close a product that did not work in a culture that prefers to quietly let things die, and how to lead a cross-functional team I had built from zero.

Mail.ru Startup Incubator was the company’s internal incubator programme. The goal was to increase the capitalisation of a five-to-six-billion-dollar company by spinning out new products. We reviewed over a hundred internal pitches and launched two new product teams. Some of the ideas were genuinely strong. None were big enough to move the valuation of a company that size. I learned that internal incubation works only when the math of the parent company allows it.

Peace.Love.Bread was a bakery and brunch café in Istanbul I helped found as investor and partner. The first truly offline business I had ever been part of. I was responsible for legal setup, premises, IT, and the website. We closed when our operating partner walked away. From Peace.Love.Bread I learned how to start a business from scratch in a foreign country, and that I need to screen co-founders and partners much more carefully.

HlamProdam was a “we sell your stuff for you” service in Moscow that I co-founded. Off-the-charts customer satisfaction, word-of-mouth growth, and an economic model that never quite worked. From HlamProdam I learned that customer love does not automatically convert into a viable business when the unit economics are too thin, and that offline operations are harder to scale, harder to control, and harder to bootstrap than pure online ones.

SyTech was a Russian software company building systems for government and private clients. They hired me as a junior Java developer while I was still at university, even though I did not know Java. I grew to senior Java developer in two years. I learned how to be a full-stack software developer, building web services across every tier from system administration to backend, frontend, and design. I learned how to maintain a codebase someone else built, and how to deliver project stages on deadline, present them to clients, and hand off the work cleanly. The full-stack skill is what allowed me to start AvtoTochki next.

ArtekNet is a Moscow internet service provider I co-founded as CTO while at university. I built the technical infrastructure: server, billing system, website, operations. We grew to 1,500 connected apartments. I learned how to make a system run cleanly enough that maintenance becomes minimal, that when no real challenge is left it is time to step away, and how to work with employees: hiring, delegating, controlling, managing people for the first time in my life.