Hi, I'm Mike 👋
A Computer Science Student building tools for visualize thinking
ML

About

Since buildspace s5, I have been working on a visual note-taking app Yboard to improve our thinking and working flow for researchers, students, and writers using AI-generated graphs. I also make cinematic music and watch movies in my spare time, loving Ghibli at the moment. If these stuffs sounds fun to you, feel free to reach out.

Skills

python
c
pytorch
langchain
mysql
cloudflare
typescript
javascript
java
react
nodedotjs
nextdotjs
postgresql
docker
git
github
visualstudiocode
figma
Projects

Check out these projects👇

I've worked on a variety of projects, mostly related to Machine Learning and Creative Tools. Here are some of my favorite projects.

Yboard

Yboard

A simple canvas/note-taking tool for branching and remixing AI-generated outputs. Instead of a linear thread, conversations on Yboard are modular and can be taken apart, branched, and combined in different ways.

Typescript
Next.js
Cloudflare
Sqlite
JustWrite

JustWrite

A text editor that allows writers to "locked in", forcing them to write continuously for a set time without switching tabs or risk losing their work. Aims to eliminate distractions and boost focused productivity.

Typescript
React
Vite
Shadcn/UI
Search Shortcuts Extension

Search Shortcuts Extension

Developed an open source Chrome extension using TypeScript and React that implements DuckDuckGo's "Bang!" customizable search shortcuts to navigate to different website quickly. Adding the !w to the search query can redirect the query to Wikipedia instead of Google

Typescript
React
Chrome
AI Storyboard

AI Storyboard

AI-generated images of scenes in a storyboard using Llama2 and a Stable Diffusion model integrated with a PDF parser to extract crucial scene descriptions and scene indexes

Streamlit
Python
Langchain
Chroma
FAISS
llama.cpp
Music Lyrics Autocomplete

Music Lyrics Autocomplete

Developed a transformers model to complete a text based on the artist's lyrics datasets scraped from Kaggle datasets. Have 6 monthly users on Streamlit community. Trained artist lyrics model includes The Beatles, Taylor Swift, and John Mayer.

Streamlit
Hugging Face
Pytorch
Kaggle
Podcast/Audio summarizations extraction

Podcast/Audio summarizations extraction

Extracting essential video information and cutting crucial video segments using ChatGPT API, Langchain output parser features, and FFmpeg.

OpenAI
Python
Langchain
FFmpeg
ChatGPT API
Pydantic
Hugging Face
Chroma
Reddit Memes Search Engine

Reddit Memes Search Engine

A meme search engine that generated a meme based on the Kaggle Reddit meme dataset using the CLIP model and semantic search.

Streamlit
Python
CLIP
Hugging Face
Reddit Kaggle Dataset
Contact

Get in Touch

Let's talk more via linkedin, twitter, and I'll respond as soon as possible.

Made with Magic UI