Loading
Kai Zhao
About
Hey guys! I'm Kai, an AI and full-stack software developer in the Bay Area studying IEOR at UC Berkeley.
I'm currently building MouseFit, a web product that helps gamers find better-fitting peripherals using hand-size, grip-style, and preference-based AI recommendations.
I'm interested in AI systems, agents, retrieval, evaluation, and building reliable software around LLMs: turning models into useful products, not just prompting them.
I build with TypeScriptPythonJavaScriptJavaC++.
Product-wise, I use React/Next.jsNode.jsPostgreSQL/SQLiteTailwind CSSGit/GitHubVercel.
AI-wise, I'm working with LLM APIsRAGEmbeddingsVector searchAgent workflows.
Hit me up on LinkedIn.
My Projects
01 // Fullstack AI Product
Mouse-Fit
A browser-based ergonomic analysis tool. Uses simple Q&A assisted with RAG/Computer Vision to analyze a real-time measurement of your hand, and matches you with the perfect gaming mouse.
02 // Chrome Extension
VentureSignal
The intelligent VC assistant. An AI-powered Chrome extension acting as a 'second brain' for investors. It parses startup websites, passively monitors for traction signals, and auto-drafts re-engagement emails.
03 // @Polylabs Inc
ScratchGPT
Debug and develop a curriculum plan around a creative agent that translates natural language instructions into functional, block-based Scratch code for K-12 students.
04 // @Someidea AI
Financial AI Agent
Designed and built the backend pipeline for a multi-agent financial analyst AI using LangChain. Use LLM to extract and summarize complex financial reports into consistent income statements and balance sheets.