🚀 Colosseum accelerator (Solana) - selected as one of ~10 projects worldwide. Grew into Tempo, which raised $250k.
Real onchain data - MEV sandwich and trading analytics for Raydium V4 and Pump.fun.
- 🚀 Colosseum accelerator (Solana) - selected as one of ~10 projects worldwide, became Tempo and raised $250k
- 🏅 Superteam Poland Grant - ~$8,500
- 🏆 Helius Redacted Hackathon - won 2 tracks
It started as a bot. A fully autonomous trading bot on Solana, written in Rust from scratch. It traded 10,000-20,000 tokens a night, running 10-12 hours completely unattended, filtering opportunities through ~15 custom filters, with sub-0.5ms execution that landed trades in the same slot as the trigger. It was profitable within the first three weeks.
The kicker: it was my first ever Rust project, built with no prior Rust or Solana experience.
To run it, we were indexing every transaction on Raydium V4 and Pump.fun into a Postgres database. At some point it clicked: we were already sitting on all this onchain data, so why not surface it. That became Sol Statz - a dashboard exposing MEV sandwich and trading activity that nobody else was showing openly.
Sol Statz was submitted to Colosseum Eternal, got picked into the accelerator as one of ~10 teams worldwide, and evolved into Tempo (raised $250k).
- Live MEV sandwich statistics across Raydium V4 and Pump.fun
- Leaderboards: top sandwichers, top traders, biggest attacks
- Per-period summaries of extracted value and attack counts
- Built on a full transaction index of both venues
- Autonomous bot trading 10,000-20,000 tokens a night, 10-12h unattended
- ~15 custom filters driving every trade decision
- < 0.5 ms execution, same-slot landing
- Profitable within the first 3 weeks
- My first Rust project - built with zero prior Rust or Solana experience
- Dashboard indexed Raydium V4 + Pump.fun in full (49,186 MEV attacks tracked in one week, Feb 2025)
| Dashboard | Attacks & charts |
|---|---|
|
 |
| Top traders | Latest MEV attacks |
|---|---|
 |
 |
- Bot: Rust, fully autonomous, tuned for sub-slot execution on Solana
- Indexer: custom pipeline ingesting all Raydium V4 + Pump.fun transactions
- Database: PostgreSQL
- Frontend: Next.js, TypeScript
Founder. I built the entire frontend, prepared the data layer that fed the indexer, and wrote the whole autonomous trading bot in Rust from scratch.
Sunset. Source code is kept private - this repo is a case study.