Summary
Overview
Work History
Education
Skills
Websites
Timeline
Generic

Harrison Li

Princeton,NJ

Summary

Self-motivated software engineer, skilled and experienced in diverse fields

Overview

2
2
years of professional experience

Work History

Software Engineer

AYR.ai
03.2022 - Current
  • Working in a fast-growing AI startup which was in the phase of developing and deploying distributed microservice applications to leverage native Kubernetes GPU cluster, and Cloud venders. As an active team member, I contributed to different aspects of software development, shown as follows.
  • Full-stack development: Developed a centralized platform management page for production environments, a real-time model training page (refactored), and data labeling tools and features. Refactor the backend logging system to manage Linux process outputs. (React.js, Django )
  • Software infrastcture enhancement: Re-architected the team's application, moving from pure source code deployment to docker and docker-compose. Optimized docker image size, Linux processes control, and deployment steps. These simplifies and stabilizes the on-premise production system. (docker, docker-compose, bash)
  • Machine Learning modeling: Proposed and integrated a transformer-based CV model to the document classification pipeline, which boosted overall accuracy to 90 percent (Pytorch)
  • Dev-Ops: Built Kubernetes GPU cluster for internal software deployment and testing; built CI pipeline for docker image build and unit test; responsible for docker image upgrade, code troubleshooting on the UAT system

Research Associate

University Of Michigan, ECE Department
06.2021 - 03.2022
  • Worked with Dr. David Garmire on a genetic data analysis Platform dev and improvement
  • Developed I/O units (Python & R) to pull large research datasets from other platforms through http requests
  • Developed Python SDKs to enable large genetic dataset mining by supporting dataset chunking, in-memory compression and decompression, chunk size adjustments, and Machine Learning algorithms iterating over chunked dataset. All codes are encapsulated in Object Oriented Design and pushed to Docker Hub.
  • Developed a NodeJS prototype application to enable running light ML model iteratively over chunked dataset and visualization of result within frontend browser

Machine Learning Intern

Diversity Policy
06.2021 - 09.2021
  • Worked on projects to research and develop deep learning models for text cleaning and keyphrase extraction to enhance the company’s web search experience.
  • Developed a BiLSTM model to classify text should be clean or not using Tensorflow, achieved 83% accuracy
  • Fine-tuned a BERT model to extract keyphrase from cleaned text using Pytorch

Education

Master of Science - Data Science

University of Michigan - Ann Arbor
Ann Arbor, MI
12.2021

Bachelor of Science - Statistics And Computer Science

University of Illinois At Urbana-Champaign
Champaign, IL
12.2019

Skills

- Full Stack (JS, Java, Python)

- Distributed system and container tech (Docker, Kubernetes)

- Database system (Postgres, MySQL, MongoDB, Neo4J, Redis)

- Machine Learning (PyTorch, Tensorflow, Sklearn)

Timeline

Software Engineer

AYR.ai
03.2022 - Current

Research Associate

University Of Michigan, ECE Department
06.2021 - 03.2022

Machine Learning Intern

Diversity Policy
06.2021 - 09.2021

Master of Science - Data Science

University of Michigan - Ann Arbor

Bachelor of Science - Statistics And Computer Science

University of Illinois At Urbana-Champaign

Similar Profiles

CREATE PROFILE
Harrison Li