Chapter 00 — First Light
EthanWang
Building systems that trade, scale, and think.
Software engineer, quantitative trader, and startup founder — walk the trail and see what grows.
Roots
Where everything takes hold.
I build things at the intersection of quantitative systems, AI infrastructure, and startups.
I'm a computer science and mathematics student at Case Western Reserve University. Currently co-founding Darch AI, where I architect high-throughput media pipelines serving 20M+ monthly impressions. Previously built AI tools at NIST. On the side, I run automated trading systems on prediction markets — reaching the top 100 on Kalshi's all-time crypto leaderboard.
The Trail
Every step leaves a mark on the path.
Darch AI
Co-Founder & Engineering Lead
Remote
- ▹Architected a hybrid microservices system (Python/FastAPI, Node.js) and high-throughput FFmpeg media pipeline to automate cross-platform content distribution, supporting 20M+ monthly impressions.
- ▹Optimized serverless resource allocation for 3,000+ monthly video jobs, maintaining 85%+ profit margins through aggressive caching and stateless execution.
- ▹Maintaining custom self-hosted infrastructure with modified authentication flows for multi-tenant B2B scheduling, while leading solution architecture for enterprise accounts.
National Institute of Standards and Technology (NIST)
Software Development Intern
Gaithersburg, MD
- ▹Built AI-driven internal tools including a help desk chatbot powered by a RAG pipeline, engineering document ingestion and optimizing retrieval strategies.
- ▹Developed and integrated a custom logging feature into an Open WebUI fork to enhance monitoring and debugging capabilities.
- ▹Collaborated with Boulder campus staff to deliver a wildfire evacuation dashboard based on stakeholder requirements.
Clearings
Things built where the light gets through.
May 2024 – Present
Quantitative Trading on Kalshi
Automated trading system executing across prediction markets with market making, momentum, arbitrage, and latency-sensitive strategies.
- ▹3,500+ monthly trades totaling 500k contracts
- ▹Low-latency infrastructure processing hundreds of GBs of market data
- ▹Predictive models using Ornstein-Uhlenbeck SDEs, Hawkes processes, and modified Black-Scholes
February 2025
Orbit Chrome Extension
Multimodal semantic search engine for saved text, audio, and video using CLIP/CLAP embeddings.
- ▹Natural language queries over captured browser content
- ▹ChromaDB integration for real-time vector similarity search
Undergrowth
The dense, living toolkit underfoot.
Languages
Frameworks
Infrastructure
Nightfall
The fireflies come out. Let's talk.
Let's build something that grows.
Always interested in new opportunities, collaborations, and interesting problems worth chasing into the dark.
ethan.wanq@gmail.com▌Designed & built by Ethan Wang — wander back to first light.