Quantitative DeFi · Risk monitoring · Cross-chain yield optimization
_{Madrid, Spain · open to remote / EU}

I founded the intelligence layer at Sail — a DeFi yield-optimization protocol on Base and Arbitrum — from a blank page, and now lead a team of two engineers building it. Three production systems, all designed and implemented end-to-end (or, in the case of the protocol, contributed to as a core author):

• Cross-chain optimization engine — constrained MDP formulation, simulated annealing with Metropolis-Hastings acceptance, CNN meta-controller predicting hyperparameters from APY/TVL history (3-5× search-space reduction). Energy-aware cost model covering execution, slippage, and TVL-dependent price impact.
• Sonar (risk monitoring) — z-score anomaly detection over multi-horizon baselines (1d / 7d / 30d), dual-mode SPIKE/TREND scoring, tier-based execution cycles, financial-inertia mechanism that cut unnecessary portfolio churn by 43 %.
• Sail Protocol (contributor) — the first onchain Separately-Managed-Account protocol for autonomous agents. A ~590-SLOC Solidity kernel mediates Safe custody, EIP-712 mandates, and gas-capped staticcall permissions. Replaces the third-party permission layer with an in-house design (audit-prep).

I also defined the database schema, built the first user-facing UIs, and represented the intelligence layer to venture investors during fundraising rounds and the CEO during roadmap discussions.

The full mathematical framework is documented in two public research papers I authored (sole author).

First six months in production: ~$700M of volume routed for ~200 users at ~$600k TVL.

All returns net of every fee in the cost model.

Sonar suspended user exposure ahead of two publicly confirmed exploits — detection happened before public confirmation in both cases:

• Moonwell oracle exploit · Feb 15, 2026 — $1.78M bad debt, 181 borrowers harmed.
• Resolv Protocol exploit · Mar 22, 2026 — ~$80M minted via a compromised AWS KMS key, ~$25M extracted.

All sole-author work.

• ✍️ Quant Mentality: How We Use Convolutional Networks to Outyield DeFi · long-form · May 2026
• 📄 Cross-Chain Optimization Engine V2 — Smart Spike Framework · paper · Jan 2026
• 📄 Cross-Chain Optimization Engine V1 · paper · Oct 2025
• 📊 Sail Agent Performance Analysis — Q4 2025 · report · Jan 2026
• 📊 Sail Agent Performance Analysis — Q1 2026 · report · Apr 2026

B.S. Computer Science — Universidad Politécnica de Madrid · 2019 – 2023 GPA 8.56 / 10. Matrículas de Honor in Operating Systems, Compilers, Semantic Web & Knowledge Graphs, Art of Programming, Prolog, Programming Project, Data Center Project.

 alonso.miguel.alvaro1@gmail.com

 alvaro-alonso-miguel