About

I'm Affaan Mustafa. SF-based builder working at the intersection of AI, finance, and developer tools. Currently co-founding Itô. An agentic desk for compute.

What I'm Building

Itô is an agentic desk for compute.

I also created Everything Claude Code, an open-source collection of skills, agents, hooks, and workflow patterns for coding agents. And AgentShield, an AI agent security scanner featured at the Cerebral Valley x Anthropic event.

Education

3 degrees by 20. BS in Mathematics-Computer Science and BA in Business Economics from UC San Diego. AA from Bellevue College. Later attended UW Applied / Computational Math grad school (MS/PhD track) and left. MIT Applied Data Science Professional Certificate.

Experience

Itô (2025 - Present)

Co-founder — An agentic desk for compute.

PMX Trade (Sep 2025 - Feb 2026)

Head of Product & Founding Engineer — Permissionless tokenized prediction markets platform, $250K+ MRR, 6-figure MAU. Built core trading infrastructure. Left due to budget cuts.

Dexploy (2025)

Creator — Instant tweet-to-coin deploys in under 300ms. Built deployment automation and infrastructure. Acquired.

elizaOS (Dec 2024 - Jan 2025)

Core developer on elizaOS, the most used AI agent framework in Web3 (17,000+ stars). Contributed to foundational infrastructure powering thousands of AI agents.

Cainam Ventures (Feb 2025 - Mar 2025)

Lead Developer — Built an autonomous DeFi AI agent swarm across aggregation, analysis, and execution. ICO-backed project reached $8M peak FDV and was featured by Forbes.

ModernStoicAI (Nov 2024 - Jan 2025)

Built the first autonomous trading agent on Solana. $2M+ in trades. Live-streamed development peaking at 70k concurrent viewers. $38M FDV.

DCUBE (Sep 2021 - Apr 2025)

Founder — Started at 18. $150K+ profit bootstrapped ecommerce business to pay tuition. Microsoft for Startups + UCSD Blackstone LaunchPad.

Research

• HyperMamba Paper — Extending Mamba with hypernetworks for meta-learning in trading. 529 views, 47 downloads.
• Discovering Market Manipulation Through Social Media Sentiment Analysis in Microcap Cryptocurrencies — Microcap crypto sentiment/manipulation research (also presented via UCSD undergraduate research channels / URC materials).
• Behavioral RL — Reinforcement learning models incorporating behavioral economics biases.
• research — Personal / UCSD / UW affiliated research code and notes.
• econ-research-wi24 / econ-research-ss24 — economics lab/project research repositories.
• quranic-arabic-recognition-dl and supply-chain-ml-optimization — deep learning and ML optimization work (including SVM / Newton's Method comparisons).

More on my SSRN profile , UW AMATH page, and UCSD Econ Research Lab.

Values

I care about depth over breadth, real iteration speed, and work that compounds. Substance over flash.