I'm Dmitriy Vasilyuk. I am a full-stack developer specializing in AI and Web3 applications.
In a previous life, I was a theoretical physics PhD student, startup co-founder, professional poker player, science and philosophy blogger, and a DeFi and algorithmic trading enthusiast.
I'm currently building DocDocGo, a multifunctional chatbot that saves you time when you have to sift through lots of websites or documents to find the information you need. It conducts online research, finds important information among hundreds of sources, and builds comprehensive knowledge bases from these sources that you can chat with. Recently featured by Streamlit. Check out the live demo and code.
In the course of working on DDG and building LLM workflows for two AI-related startups, I have consolidated the reusable tools I've developed into LightLang, a lightweight Python package that makes working with LLMs easier without adding opaqueness and too much abstraction (looking at you, Langchain 😉). Feel free to give it a try, and you can find some examples in the LightLang Cookbook repo, which you can run without an API key (you heard it right!).
Bet Czar dApp - Manage and enforce bets on EVM blockchains. The smart contract is deployed on Goerli. I built this dApp when I was learning Solidity and React. It's a fun way to make bets with friends and enforce them with smart contracts.
VocabMeThis - Measure and improve your vocabulary. This Flask app uses a frequency dictionary built from NLTK corpora to rank words by difficulty and suggests words at just the right difficulty level. Great for English learners.
I'm passionate about physics and analytic philosophy. Here are some articles from my ReasonMeThis blog and StackExchange contributions: