
Building "Impossible" MEV Technologies to Successful Exit
Founding Engineer @ Staging Labs
Built ultra-low-latency MEV systems that intercept malicious transactions before they confirm, later adapted for enterprise compliance and custody, leading to the acquisition by Merkle Science.
1The Challenge
Web3 security had a blind spot.
Once a malicious transaction hit the mempool, users were effectively helpless. Wallets could be drained in seconds, and most "security" tools only told you after the damage was done. Even industry veterans believed reversing or neutralizing transactions wasn't possible.
Our goal was simple to describe and extremely hard to execute: beat attackers inside the mempool, at speed.
2The Solution
I joined as a founding engineer to help turn an early proof-of-concept into a production system under a tight deadline.
We built a mempool-monitoring and frontrunning system that could detect malicious behavior in real time and submit a higher-priority counter-transaction before the attacker's transaction finalized.
The core mechanism worked like this:
- •Users set explicit spend limits and safeguards
- •Assets were locked behind a minimal, audited contract that could only migrate funds to a user-owned cold wallet
- •Our system streamed mempool data, decoded transactions in real time, and classified intent
- •If a safeguard was violated, a counter-transaction was constructed and broadcast with a slightly higher priority fee, executing before the attacker's transaction confirmed
In practice, attackers saw failed transactions while funds were safely migrated out of reach.
After launch, I focused on performance. The original system executed in ~1s end-to-end. That wasn't good enough. I reworked the execution path, networking, and decoding pipeline to push latency down to ~60ms end-to-end, with ~3ms of actual processing time, dramatically increasing coverage during late-block scenarios.
We later extended the same MEV-based techniques beyond wallets:
- •Support for non-fungible tokens (NFTs) and other token-based assets
- •Enterprise-grade transaction mitigation
- •MPC and multisig compatibility
- •A compliance "vault" model that quarantined suspicious funds within the same block
This ultimately led to live demos with major financial institutions, including work connected to BNY Mellon via Chainalysis.
3Results
- 99.7% reduction in execution time (≈1s to ≈3ms processing)
- 3 worlds-first MEV-based security systems designed and deployed
- Production mainnet deployments, contracts audited
- Enterprise PoCs delivered to global financial institutions
- Acquired by Merkle Science (June 2024)
