← Back to home

RÉSUMÉ / OCTOBER 2026

Haiyang Yu

AI product builder and researcher with Berkeley data science and statistics training and a UCLA Master of Financial Engineering. Turns firsthand user insight and research questions into product decisions, and directs AI-assisted implementation of working tools.

EXPERIENCE

Independent AI Research & Product Practice

Mar 2026 - Present

Product Lead and Research Author

  • Multi-agent coordination · pneu. GitHub · Made separate coding-agent sessions work together with their native tools and approvals: conceived an open-source tool for exchanging tasks and reviews across harnesses; directed AI-assisted implementation; submitted to OpenAI Build Week.
  • Product development. Took pneu from concept into daily use, defining workflow requirements, setting priorities, and directing AI-assisted implementation; continually refined the coordination architecture and user experience through firsthand use.
  • Delivery rules. Specified delivery rules that keep ordinary messages pending while a human task is busy and prevent blind retries when admission is unconfirmed.
  • Research data tools. Directed the design of a research lake combining market data, SEC filings, and macro series; defined how economic dates differ from retrieval dates and how filings become dated holdings and share-count series that AI agents query through MCP.
  • Digital-asset treasury research. Sole-authored a working paper deriving when maximizing Bitcoin per share diverges from shareholder value, submitted to AFA 2027. In related capital-stack research, framed treasury-company equity as a perpetual option on Bitcoin holdings and designed its calibration to Bitcoin futures, ETF implied volatility, and the Treasury curve.

Nora

Sep 2025 - Mar 2026

Founding Developer Relations Engineer · AI coding agent for on-chain development · Product Hunt #1 Product of the Week

  • Product direction. Continually helped shape Nora's product roadmap through competitive research and developer insights; recommended an on-chain focus that the CEO adopted.
  • Vendor decision. Settled an LLM-gateway choice with evidence: designed a latency and reliability comparison against OpenRouter, separated spending-cap failures from provider failures, and declined the switch.
  • Developer demos & education. Directed working demos of a Solana order book and an EVM prediction market; designed OpenBuild courses and Solana tutorials on agent planning, testing, and replanning, with 1,000+ participants across the series, 300+ of whom joined an ongoing developer community.
  • Content and community. Gave Nora a story of its own, beyond the models it supported: owned content and release communications across X, Discord, and WeChat for 12 product versions, framing model-support releases around Nora's reasoning layer; company posts drew ~125K views.
  • Ecosystem engagement. Spoke on an Odaily Solana X Space (1,000+ listeners), designed and delivered a remote workshop for Monad Chengdu Blitz and organized the side bounty; the hackathon had 50+ participants and the bounty received 6 submissions. Designed a “From Prompt to Protocol” workshop for Crecimiento and wrote DeFi insights for BNB Chain × YZi Labs Hack Abu Dhabi.

QC Capital and Consulting

May 2025 - Sep 2025

Summer Investment Analyst · Multi-family office

  • Investment theses. Argued that US stablecoin adoption would arrive as a back-end settlement change, with value accruing beyond issuers to the infrastructure and integration layer; developed this and a thesis on prediction markets as information pricing for partners.
  • Polymarket memo. Authored the Polymarket investment memo and market-sizing deck, arguing it is an information market rather than a betting company; the office then invested in Polymarket through a venture fund.
  • Bitcoin research. Authored memos on how US Bitcoin miners turn holdings into cash and on Bitcoin's rolling correlation with equities, gold, and Treasury yields; the work led to the office's investment in a Bitcoin mining facility.
  • AI research workflow. Designed a four-stage LLM research workflow for portfolio reviews and macro briefs; produced five monthly reviews across six sectors and eight macro briefs for partners.

Project Twelve

Nov 2021 - Jun 2022

Founders' Associate Intern · Web3 gaming studio

  • Product analysis. Analyzed StarShark user activity and transactions with a self-built Dune SQL dashboard, identifying a sharp decline after a change to the game's reward algorithm; concluded that engagement depended on token incentives, and tradable on-chain assets could not make up for weak gameplay.

H Capital

Mar 2021 - Jun 2021

Investment Intern · Growth equity

  • Research origination & leadership. Originated a field study of Luckin Coffee's recovery, secured Investment Committee backing, and independently led 80 part-time observers across 25 cities.
  • Research design. Turned the recovery thesis into store-traffic and pricing tests; designed the store sample, collected a 1,000-order price sample, and wrote 15+ expert-interview notes for the team's unit-economics model.
  • Financial analysis. Estimated store-level unit economics and sensitivities in Excel from field observations and expert interviews.

SELECTED PROJECTS

TALKS & WORKSHOPS

EDUCATION

UCLA Anderson School of Management

Dec 2025

Master of Financial Engineering

Coursework: Econometrics, Empirical Methods in Finance, Machine Learning, Computational Methods, Stochastic Calculus, Derivatives, Fixed Income.

University of California, Berkeley

Dec 2023

B.A., Data Science & Statistics

Coursework: Data, Inference & Decisions; Principles & Techniques of Data Science; Data Mining & Analytics; Machine Learning; Data Structures; Time Series; Stochastic Processes.

CAPABILITIES

Python, SQL | Claude Code, Codex, MCP, git | English & Mandarin bilingual | Cantonese intermediate

My work spans product direction, research, and writing. Software projects are developed with AI coding agents.