# About me (/about-me) I'm Oleg Polakow, the creator of [VectorBT](https://github.com/polakowo/vectorbt) and VectorBT PRO. I build quantitative software for a living and have worked in software for more than a decade. These days, most of that time goes into giving independent developers and traders better ways to test ideas, understand the results, and stay in control of their research. ## A little background \[#a-little-background] ![Oleg Polakow standing by the sea](/assets/images/about/oleg-polakow-travel.webp) I'm from Lviv, Ukraine. As a teenager, I joined Ukraine's Minor Academy of Sciences, where I assisted a professor and learned about quantum mechanics and astrophysics. I then studied physics at the University of Lviv for a year. Back then, I hoped to become a physicist. As part of my studies, I wrote a program in Pascal that calculated astronomical events. That was when I discovered that I enjoyed writing code even more than studying physics. At seventeen, I moved to Berlin alone and started over in a new country. I began studying Computer Science at TU Berlin. I later moved to Munich, switched to Information Systems with a stronger focus on business, and earned my master's degree from the Technical University of Munich. It wasn't an especially linear route, and I learned a lot as I went. Somewhere along the way, I also earned a black belt in Karate-Do. Science, software, and Karate may look unrelated, but all three reward curiosity, patience, and the ability to fail repeatedly without taking it personally. ## Before VectorBT \[#before-vectorbt] My education combined physics, computer science, software engineering, data, and business. I studied at the Technical University of Munich and spent an exchange year at the Technical University of Denmark. I was especially drawn to machine learning and big data. I liked taking a pile of numbers and finding out what it was actually trying to tell me. I learned most by building. My early work ranged from fraud detection with SAP HANA and research software for analyzing digital transformation to an iOS and server-side product with ZEISS and Python analytics at Peergrade. The projects were different, but each helped me discover how much I enjoyed building software around practical problems. ## How VectorBT began \[#how-vectorbt-began] I started building VectorBT in 2017 while working on a crypto monitoring system that followed hundreds of coins. I needed a faster and more flexible way to explore the data and test many ideas. Instead of running one strategy configuration after another, I represented many configurations as arrays and evaluated them together. That simple shift became the foundation of the project. I released VectorBT publicly in 2020 and have kept evolving it ever since. What began as a tool for high-speed parameter exploration grew into an open-source library, a professional research platform, and a community around both. The original architecture is still at the center of it all. VectorBT was supposed to solve one research problem. It responded by becoming my full-time job. ## How I think about research \[#how-i-think-about-research] VectorBT was built for exploring many possibilities, but I don't believe research should end by selecting the single "best" result. That is often a shortcut to overfitting. I would rather inspect the whole parameter surface, test different assets and periods, and understand why a strategy behaves the way it does. Speed matters because it gives us more room to ask better questions. I also prefer building blocks over black boxes. Data, indicators, signals, simulation, and analysis should be understandable on their own and flexible enough to combine in unusual ways. Many useful improvements began with a user showing me a problem I hadn't seen before. Real examples often make it clear what VectorBT needs next. ## How I build software \[#how-i-build-software] Software has never been easier to start building, and I like how many new ideas that makes possible. My own focus is on what happens after the prototype: making a system correct, coherent, and maintainable after years of growth. I stay close to the code and the people using it. I personally review, test, and revise every significant change. Sometimes I spend an unreasonable amount of time on something most users will never notice. Ideally, that is why they never notice it! ## Beyond the code \[#beyond-the-code] Away from work, I enjoy traveling, discovering new cultures, and practicing Shotokan Karate-Do. I tend to like activities where progress comes from curiosity, repetition, and occasionally being reminded that I know less than I thought. That may also explain why I still enjoy software. You can find my work and writing on [GitHub](https://github.com/polakowo) and [Medium](https://polakowo.medium.com/), connect with me on [LinkedIn](https://www.linkedin.com/in/polakowo/), or contact me at .