Hey everyone,
I’m currently building a flash loan arbitrage bot focused on DeFi protocols like Uniswap, and I’m looking for knowledgeable people to collaborate with and exchange ideas.
The core strategy follows the classic pattern:
Flash Loan → Mint → Swap → Swap → Swap → Burn → Repay Loan
But the real challenge lies in efficiently detecting profitable cycles and accurately modeling the economics — especially around Uniswap’s pricing formulas (e.g., x * y = k invariant), fee structures, slippage, and gas costs.
What I’m working with:
- Backend in Rust and Go (Preformatted textclean, modular design)
- Real-time price fetching and pool state tracking
- Flash loan integration (Aave, Balancer, etc.)
- Graph representation of liquidity pools (tokens as nodes, pools as edges)
Where I need help:
- Mathematical modeling – refining the profitability formula considering fees, reserves, and multi-hop impacts.
- Graph algorithms – efficient cycle detection (especially negative-weight cycles via Bellman-Ford or Johnson’s algorithm), but optimized for speed and scalability.
- Uniswap V2/V3 mechanics – deeper insight into tick math, concentrated liquidity, and accurate swap simulations.
I’ve got the infrastructure and basic logic running, but I want to improve accuracy and reduce false positives in opportunity detection. The goal is not just a proof-of-concept, but something robust and production-ready.
Looking for:
- People with experience in DeFi arbitrage, graph theory, or quantitative trading systems
- Interest in collaborative development — sharing knowledge, reviewing logic, optimizing code
- Open to working together long-term; could evolve into a shared project or even startup down the line
If you’re into crypto, math, and low-latency systems — I’d love to chat. Even if you’re building something similar, let’s talk! There’s more than enough space for learning from each other.
Feel free to reply here or DM me. Let’s build something smart together.