A fireside panel on stage before a packed auditorium at Columbia Business School

Systems thinking
for applied AI

A curious human exploring how frontier models translate into systems institutions can actually deploy. Always learning, experimenting, and occasionally sharing agent-native workflow designs.

Explore projects Read my writing

About

Taewan Yoon, CFA

LinkedIn GitHub

Columbia MBA/MS. Built and scaled Claude's NYC builder ecosystem from 0 to 1,000+ members. Led the AI Club through 4x growth to 900+ paid members, and co-organized the flagship AI Startup Challenge. Founded BRAIN, a cross-school startup network tackling the hardest business challenges with frontier AI.

AI strategist, product manager, and startup community builder. Shipped agentic workflows as Manus AI Fellow, and co-founded AI-native fintech startup (acquired by a major insurance firm). Mentor at CFA Society New York.

Projects

Featured Claude Code 2026·06·28 claude-ensemble: Achieving Frontier-Level Performance with Claude Code A Claude Code kit for the most challenging tasks. Runs a Sonnet-5 panel with a high-effort Opus judge and a code-grounded verify-loop, entirely on a Claude subscription (no API key). Backed by blind, length-controlled A/B evals.

Featured RAG 2025·09·30 RAG Pipeline for Quantitative Investment Research Built a LightRAG pipeline that combines a knowledge graph with vector embeddings to make 4,000+ academic papers queryable, surfacing contrarian investment strategies with 2.4x more comprehensive responses than naive RAG. Presented at MathWorks.

Product Management 2026·05·08 Access to Experts — CNN Product Strategy A 12-week product strategy engagement with CNN: diagnosed why engaged readers leak attention to Substack, podcasts, and X during breaking stories, then designed a credentialed expert reaction layer and stage-gated rollout endorsed by senior product leadership.

Causal Inference 2026·03·11 Causal Inference — Interactive Study Guide A single-page interactive web app covering causal inference from potential outcomes to Double ML, with canvas visualizations, KaTeX formulas, and a full method comparison framework.

4 case studies All projects

Writings & Insights

Featured Claude Code 2026·08·04 I Deleted Half My Agent's Instructions, and Here's What Happened 729 blind trials on CLAUDE.md and skill files. The same document scored 87% and 13% depending on nothing but the filesystem underneath it.

Featured Developer Marketing 2026·05·14 Let the Flywheel Spin on Top of the Translation Layer How 6+ years of building, shipping, and 40+ AI rooms came together to uncover a truth about earning trust in AI

Claude Code 2026·09·01 Who Holds the Loop I spent a week running GPT models inside Claude Code. Here is the harness that decides everything, OpenAI's new plugin, and the receipts from every claim I tested.

Knowledge Management 2026·07·21 What Form Should Knowledge Take? The evidence from AI frontier settling on what curated knowledge should look like as a system. And a new place for RAG and retrieval turns out to serve it.

Knowledge Management 2026·06·21 How I Think About Knowledge in Finance Finance spent two decades improving its data systems. It did far less to preserve the reasoning behind decisions. AI has lowered the cost of finding that reasoning. This essay explains how financial firms can manage it.

Contact

Let's connect, build, and inspire.

Always learning, always building agentic systems and frameworks. If you're doing the same, I'd love to connect.

[email protected] LinkedIn GitHub