Raven Trading

Raven Trading is a proprietary crypto market making and high-frequency trading firm, founded in January 2023 by three engineers from Wintermute’s DeFi team. It is run by CEO Petar Kostov and CTO Yuriy Myronovych, who built and led the DeFi desk at Wintermute before co-founding Raven. The firm is headquartered in Sofia, Bulgaria, and also has an office in Milan, Italy.

Raven trades mainly for its own account, algorithmically and around the clock across the US, Asia-Pacific and European sessions, quoting two-sided prices on centralised exchanges, DeFi venues and prediction markets. It also provides liquidity to exchanges and token projects under service level agreements, and helps new venues get started: it works with them on technical setup and market design in the weeks before launch and supplies the opening liquidity on launch day. Since moving into prediction markets in the second quarter of 2025, it has quoted more than 3,000 event contracts on more than ten venues, including Kalshi and Polymarket.

Hack VC led Raven’s 2.7 million US dollar seed round in 2024, with Wintermute Ventures taking part. In October 2026 the firm raised a strategic round from CMCC Global and Coinbase Ventures at a pre-money valuation of 90 million US dollars, to grow its prediction market business, and the same month EDX Markets announced Raven as a liquidity partner on its institutional crypto exchange. The trading systems run on Raven’s own low-latency infrastructure, and the stack is written in Rust.

What Raven Trades

• Centralised exchanges
  Two-sided quotes in spot and perpetual futures on venues such as Binance, Coinbase, OKX, Bitget and Gemini.
• DeFi venues
  Market making on on-chain order books, RFQ systems and automated market makers, including Hyperliquid.
• Prediction markets
  Event contracts on politics, crypto, macro, sports, commodities, entertainment and weather. Crypto contracts are priced off the options volatility surface and hedged in perpetuals and spot, and separate trader roles cover crypto and sports markets.
• Token and launch liquidity
  Liquidity programmes for token issuers, market making at token generation events, and opening liquidity for new exchanges and prediction market launches.

Culture and Benefits

Raven is a young firm built by engineers, and half of its open roles are in software and network engineering. Its careers page names three values: ownership, meaning you fix the problems you find and build the ideas you have; problem solving; and competitiveness, with the aim of getting sharper and quicker every day. All roles are in the office in Sofia or Milan, with no preference between the two, and none are remote. The job ads do not ask for a degree or certifications, only relevant experience and fluent English. Raven does not publish a benefits package, so ask about pay and relocation in the intro call.

Hiring at Raven

Openings cover software engineering for exchange integrations, market data and the trading platform, network engineering for the trading infrastructure, DeFi algorithmic trading, and quantitative trading in crypto and sports prediction markets. You apply through the role pages on the careers site, and the firm also welcomes a short note at noholiday@raven-trading.com if no listed role fits.

The selection process usually runs in four stages:

• Intro call
  A short conversation with the hiring lead.
• Technical deep dive
  A technical interview specific to the role.
• Take-home or case
  A scoped problem based on the work the team actually does.
• Final round
  An onsite or final round where you meet the wider team.

The quant trader roles ask for real derivatives experience: volatility surfaces, greeks, and pricing digital and barrier options off the crypto volatility surface, with the delta hedged in perpetuals and spot. They also ask for judgement on market microstructure, such as queue position and adverse selection, strong applied probability and statistics, and clean backtesting, so expect the technical deep dive and the case to test those. The engineering roles ask for Rust or modern C++ and experience running real-time production systems, and you should be ready to explain how you measured a system before you optimised it.