Work
OmniOS OÜ (opens in a new tab)
Founder and lead engineer
OmniOS OÜ is a software company in Tallinn. I started Vultax in 2024 as an independent project and founded the company around it in September 2026. It now builds two products, OmniOS and Vultax, on one rule: every number shows where it came from and when.
It is also an experimental research company for novel uses of AI agents: software that plans and carries out multi-step work on its own. What we learn goes into both products.
My wife, Sophia Kurg, joined in September 2026 as business advisor. She runs research partnerships and outreach for both products, and brings a background in financial-crime compliance and a law degree from TalTech.
OmniOS (opens in a new tab)
Estonian companies, answered from public records
OmniOS answers questions about Estonian companies. You ask Vi, its assistant, and Vi reads the e-Business Register, the Tax and Customs Board and the Public Procurement Register, links the record behind every figure, and shows where the data stops. It has been in free public beta since 23 September 2026.
The same records feed a view of the Estonian economy and our own forecasts of official releases. Each forecast is published before the figure and stays on the record afterwards, hit or miss.
Vultax (opens in a new tab)
Research for prediction markets and crypto
Vultax began as a crypto terminal: prices, order books, whale activity and cross-exchange comparison across twenty-four exchanges. It has grown into a research workspace that covers prediction markets too. You can follow a market from its price to the people, liquidity and events behind it: live Polymarket prices, analytics on 80,000 traders' wallets, a whale tracker with dated evidence, event monitors and a paper workspace for simulated positions.
The live side runs in the Terminal. The research is published as studies that state their data and method.
The part I find most interesting has not changed. Detecting wash trading, for example, means reading several behaviours together: circular value movement, timing regularity, volume, price impact and the relationships between transactions. Writing a rule that flags something unusual is the easy half. The real problem is working out how much evidence each observation provides, and combining what is available without manufacturing certainty the data does not support.
Genius Sports
Football trader, May to July 2024
Priced live football markets in real time, across simultaneous fixtures: independent judgement calls made in seconds and defended afterwards, where a missed detail is immediately visible and immediately costly. Much of it was watching for price movement that did not match what was actually happening in the match.
Writing
Background
Bachelor of Engineering in IT Systems Development at Tallinn University of Technology, expected 2027.
I mostly work in Python, TypeScript and SQL: FastAPI, PostgreSQL and TimescaleDB, Redis, ClickHouse and Next.js, deployed on Linux with Docker, nginx and Cloudflare.
I am most interested in mathematics and statistics as engineering tools rather than abstract exercises: probability, distributions, optimisation and time-series behaviour.