Alex Chan

About Me

Mission

My goal in life is to spread happiness and excitement, whether it's through sharing my passions and interests or in helping others that live in less fortunate circumstances. My belief is that there is no border between doing what you love and the work that you do, and that applying your skills to what you care most about is the most effective way of living life.

Bio
Right now I am currently a senior at UC Berkeley studying computer science. My interests are centered around machine learning and deep learning, particularly in natural language processing.

One of my favorite things to do is build and get people excited about tech. Last year I lead various engineering projects for UAVs@Berkeley, most notably FlyWave, a drone that can be autonomously controlled by detecting human arm gestures. For VR@Berkeley, I am working on SculptML, an educational AR visualization app for neural networks. I also work on the outreach team, where I have coordinated demos of VR startups and the work that VR@Berkeley does.

When I'm not building neural networks, I produce electronic dance music under the name Hydratone. My dream is to find ways to integrate machine learning to solve problems in audio synthesis, and find ways to streamline the music production process.

Other interests of mine include reading, playing basketball, cooking, listening to hip hop and electronic dance music, hiking and jamming on the guitar.

SoundCloud

Experience

HubSpot

Machine Learning Models Intern

I deployed machine learning models to the HubSpot product, as well as developed research models oriented towards new product features. Currently in the SalesHub is a topic level sentiment analysis model I developed that had a 28% improvement in AUROC. I also developed a suite of models for message recommendation, based on the 2016 Google Smart Reply paper. I retrieved the topics for different call transcripts using Latent Dirichlet Analysis.
CellScope, Fletcher Lab

Research Intern

My role was to to find ways to streamline the process of analyzing blood cell samples using Cellscope's microscope technology. During my time there, I trained a convolutional neural network to identify the number of malaria eggs in a cell image. I wrote code for image and video data processing and augmentation for training data. I also created a web application for researchers to view image and video data of blood cell samples.
UAVs@Berkeley

Project Manager

I founded the R&D Software Division of UAVs@Berkeley, with the aim of integrating state of the art machine learning and deep learning models for autonomous drone tasks. I was the project manager for the first projects of this division. Most recently, I led the FlyWave project, which was designed to autonomously manuever a drone by detecting arm gestures with computer vision and deep learning.
Berkeley Institute for Data Science

Modules Developer

I wrote iPython Notebooks on using the HathiTrust’s HTRC API to process text data for natural language processing tasks. My work is taught in Berkeley’s L&S 88-5 course. I presented my tutorials at HTRC UnCamp 2018, a conference for digital text research sponsored by HathiTrust.
Commvault

Software Engineering Intern

I created the Commvault Video Search Engine, a video player application that can search for videos based on their audio transcriptions. The engine implements Microsoft Azure Media Indexer to transcribe audio content and Commvault Analytics Engine for databasing and query searches. It is currently deployed to Commvault's internal services for searching for training videos and archived talks and presentations.
UC Berkeley

Lab Assistant for CS61A

I tutored students during labs and office hours for CS61A (Structure & Interpretation of Computer Programs), the introduction computer science class at Berkeley. Topics include: inheritance, recursion, trees, linked lists, and streams. Taught in Python, SQL, and Scheme.