Dez Calimese
“

Agents, protocols, and the infrastructure between them.

Applied AI and blockchain engineer in New York City. Five years taking products from zero to one: multi-chain AI agents secured with threshold cryptography, ERC-4626 vaults, and security-reviewed DeFi integrations across EVM and Solana.

Dez Calimese

  • Agent-driven products
  • MCP & AI integrations
  • Smart contracts & vaults
  • Security audits
  • DeFi protocols — EVM & Solana

About

Five years on founding teams, building AI agents, DeFi infrastructure, and the security work that keeps them safe.

Now

Co-founder of Quincy Labs, an AI and blockchain research studio. I built Sui Sec Bench and Solana Sec Bench, AI security benchmarks for Sui and Solana, build MCP integrations over an 8-year research base, and lead UI/UX on Emergency Passport, an AI-native app surfacing critical patient context during sickle cell crises.

Before

Founding Engineer at Omo Protocol, leading smart contracts, frontend, and AI agent integrations. This work helped secure over $1M in pre-seed. Before that, Lead Frontend Developer at Bricks Exchange, a real estate tokenization market.

Security

Completed the Rektoff × Solana Foundation Rust Security Bootcamp, adding formal Solana/Anchor audit methodology to a background in smart-contract security. My capstone audit of a Solana lending protocol found multiple high-severity bugs, each with a working exploit. Most recently I built Sui Sec Bench and Solana Sec Bench to test whether AI models can find, fix, and prove bugs in Sui Move and Solana programs.

Otherwise

Outside of engineering I analyze DeFi markets, make music, and explore audioreactive and 3D art.

Work (09)

  1. Neko DeFi screenshot

    Neko DeFi

    DeFi protocol interface featuring token swaps, perpetual trading, vault deposits, and portfolio tracking. Modern dark UI with real-time on-chain data.

    React / Next.js / TypeScript / Web3

  2. Bricks Exchange MVP screenshot

    Bricks Exchange MVP

    Decentralized exchange UI for swapping USDC for BRX, staking tokens, and managing balances on-chain.

    Next.js / Wagmi / Viem / Tailwind

  3. RSVP dApp screenshot

    RSVP dApp

    Decentralized event management app for creating and RSVPing to events fully on-chain.

    React / Next.js / Wagmi / Ethers

  4. Minting Melodies screenshot

    Minting Melodies

    NFT marketplace for artists to sell downloadable song collections as digital collectibles, with automated royalty splits.

    React / Next.js / Redux / Thirdweb

Security Research

Nov 2026 — Quincy Labs, first baseline

Solana Sec Bench

An open benchmark that tests AI models on finding, fixing and proving vulnerabilities in Solana programs (Anchor and Pinocchio). Every answer is graded by running code in LiteSVM, not by a judge model. A fix must build, pass functional tests and block a withheld reference exploit. A proof must pass on the vulnerable program and fail on the fixed one. Across 5 local open-weight models and 3 cases (one clean control): Find 31/45, Fix 9/30, Prove 0/30 first try and 8/30 with test feedback. Finding, fixing and proving are different skills: the best finder (gpt-oss-20b) proved nothing.

Solana / Rust / Anchor / Pinocchio / LiteSVM / Python / LLM Evaluation

View results ↗Repository ↗Sui Sec Bench ↓

Small pilot, not a ranking.

2026 — Quincy Labs, v1 pilot

Sui Sec Bench

An AI security benchmark for Sui Move smart contracts. LLMs must detect the vulnerable function and bug class, patch it so the fix builds, passes functional tests, and blocks the exploit, and write an exploit test that passes on vulnerable code and fails on fixed code. Across 6 local models: right function 34/36, correct patch 1/36, assisted exploit test 10/12. Models locate bugs but rarely fix them.

Sui Move / Python / LLM Evaluation / LM Studio

View results ↗Solana Sec Bench ↓

Small pilot, not a ranking.

2026 — Rektoff × Solana Foundation Bootcamp — Capstone

MetaLend Security Audit

Audited a Solana lending protocol and surfaced multiple high-severity vulnerabilities, each delivered with a working Anchor proof-of-concept confirming exploitability.

Solana / Anchor / Rust / DeFi Security / PoC

Full report available on request.

Toolkit

AI Tools

  • Claude Code
  • Codex
  • Cursor
  • Muse
  • Grok

Agents & Protocols

  • MCP
  • LangChain
  • ElizaOS
  • Hermes Agent
  • OpenClaw

Smart Contracts & Security

  • Solidity
  • Foundry
  • Slither
  • Echidna
  • Mythril
  • Anchor
  • Fuzzing / Invariants
  • MEV Protection
  • Threat Modeling
  • Gas Optimization

Languages

  • TypeScript
  • Rust
  • Python
  • Solidity
  • Move
  • Cairo
  • SQL

Multi-Chain Infrastructure

  • Lit Protocol
  • Threshold Crypto
  • Cross-Chain Messaging
  • ERC-4626 Vaults
  • Hyperliquid
  • Morpho
  • Uniswap
  • Subgraphs
  • Tenderly

Frontend & Tooling

  • Next.js
  • React
  • React Native
  • Wagmi
  • Viem
  • Tailwind
  • Docker
  • AWS
  • CI/CD

Experience

  1. 2018 — Present

    Quincy Labs

    Co-Founder

    Co-founded an AI and blockchain research collective and studio building agent-driven products across DeFi, cryptography, and applied AI. Built Sui Sec Bench and Solana Sec Bench, AI security benchmarks that test models on finding, fixing, and proving bugs in Sui Move and Solana programs, on sec-bench-core, a shared harness I built for both. Built MCP integrations surfacing insights from a 500+ channel, 8-year research base to power agentic workflows, and maintain deployment infrastructure across the collective (Docker, AWS, CI/CD). Contributing engineering and leading UI/UX on Emergency Passport (Red Cell Systems), an AI-native app surfacing critical patient context during sickle cell crises.

  2. 2023 — 2025

    Omo Protocol

    Founding Engineer

    Architected ERC-4626 standardized vaults in Solidity + Foundry and a spot/perpetuals trading platform on Hyperliquid. Built a multi-chain AI agent stack (ElizaOS Lit plugin, Coinbase MPC wallets, Nucleus SDK). Tenderly virtual testnets lifted testing efficiency 50% — work that helped secure $1M+ in pre-seed.

  3. 2022 — 2024

    Bricks Exchange

    Lead Frontend Developer

    Principal engineer over a four-person frontend team. Onboarded and mentored three junior devs through code review and pairing, cutting delivery timelines 25%. Shipped production dApps with Next.js, Wagmi, Viem, and TypeScript.

  4. 2022 — 2023

    Minting Melodies

    Chief Technology Officer

    Owned end-to-end blockchain delivery for an NFT music platform. Automated 80% of royalty distribution via smart contracts and stood up test-automation pipelines for reliable releases.

  5. 2026Education

    Rektoff × Solana Foundation

    Rust Security Bootcamp

    Completed an intensive Solana/Anchor audit program, adding formal audit methodology to a foundation in smart-contract security and full-stack engineering. Capstone: the MetaLend security audit.

    Previously

    RareSkills Advanced Solidity 2025–26 · Encode Club Expert Solidity 2023 · Encode Club Solidity 2023 · Ackee Winter School of Solana 2023 · Encode Club Algorand Camp 2022–23 · Encode Club Cairo Camp 2022 · StarkNet Basecamp 2022 · Consensys Academy Ethereum Development 2021–22 · Secureum Smart Contract Auditing 2021 · Nucamp Python, SQL & DevOps 2021

Open Source

Tools (03)

  1. Quincy-Labs/solana-sec-bench

    Solana Sec Bench

    Benchmark cases and graders for finding, fixing and proving vulnerabilities in Anchor and Pinocchio programs. Fixes and proofs run in LiteSVM against a withheld reference exploit.

    Solana / Rust / LiteSVM / Python / LLM Evaluation

    View repository ↗
  2. Quincy-Labs/sec-bench-core

    Sec Bench Core

    Shared harness behind the Sui and Solana benchmarks: model adapters, run loop, grading, leak check and leaderboard.

    Python / LLM Evaluation

    View repository ↗
  3. Quincy-Labs/sui-sec-bench

    Sui Sec Bench

    Python harness that runs local models through LM Studio and grades detect, patch, and exploit tasks automatically with the Sui Move compiler and tests, plus an agentic exploit mode with up to 5 iterations. Public grades and cases, with exploit tests kept in a restricted bundle.

    Python / Sui Move / LM Studio / LLM Evaluation

    View repository ↗

Contributions (01)

  1. elizaOS/eliza — PR #2703 — Merged

    Lit Protocol Plugin

    Multi-chain transactions plugin leveraging Lit Protocol's Programmable Key Pairs (PKPs) and threshold cryptography — enabling secure, decentralized key management for AI agents across EVM and Solana. Built for Omo's agent stack and merged into the open-source ElizaOS framework.

    TypeScript / Lit Protocol / EVM / Solana / Threshold Crypto

    View pull request ↗

Let's buildsomething

Write directly

dezcalimese@gmail.com

Or leave a note