# Mingzhe Hu

> SoFi

Location: United States, United States
Profile: https://flows.cv/mingzhe

Software engineering in Fintech.

## Work Experience
### Software Engineer, Loans Platform @ SoFi
Jan 2023 – Present | San Francisco, California, United States
N times exceptional performance rating but no further appreciation from senior leadership

Past internal tech talk topics: 
* Snowflake Kafka connectors
* Java to Kotlin migration
* Modern Airflow best practice

Responsibility: Working for the team of loans platform, with a concentration on data processing and AI-driven self-service portal. Primarily focusing on full-stack development and data pipeline, using PostgreSQL/MongoDB as db, Kafka as streaming, dbt-core as modeling, Snowflake as a cloud warehouse, gradle-plugin flyway as db migration, AWS Airflow as a data orchestration tool, Monte Carlo & Datadog as monitoring tools, Kotlin/Springboot as microservices, and React.js/MUI as UI.

### Associate @ Columbia University Department of Computer Science
Jan 2023 – Jan 2023 | New York City Metropolitan Area
[2023 Spring] Associate II of ELEN 6889 Large Scale Stream Processing under the supervision of Deepak

### Graduate Teaching Assistant @ Columbia University Department of Computer Science
Jan 2022 – Jan 2022 | New York City Metropolitan Area
[2022 Fall] Teaching assistant of COMS 4995 deep learning for computer vision course under the supervision of Prof. Peter N. Belhumeur
[2022 Spring] Serve as a course assistant of COMS 4732 under the supervision of Prof. Carl Vondrick in 2022 spring

### Software Engineer Intern @ AI Model Share
Jan 2022 – Jan 2022 | New York City Metropolitan Area
AutoML along with auto data preprocessing and transforming everything into a pipeline, deployed and dockerize model online via microservices with Amazon Web Service Lambda as a S3 proxy in API gateway

### Software Intern - Applied Research @ NVIDIA
Jan 2022 – Jan 2022 | Santa Clara County, California, United States
Working in the Metropolis team, with a focus on multi-target multi-camera tracking (MTMC) streaming data pipeline with Apache Kafka. We primarily worked with an in-house dataset for people tracking, which was used as the problemset in 2023 AI City Challenge. We set up a baseline for people tracking across the cameras with multi-dimensional associations. This can be found on NVIDIA GTC summit 2022 special track presentation.

### Graduate Research Assistant with Prof. Zoran Kostic @ Columbia Engineering
Jan 2022 – Jan 2022 | New York, United States
Works in the smart city traffic group, COSMOS lab, focusing on multi-camera re-identification with unsupervised learning in both USL and UDA.
Detection model: tuned YoloV4
Re-identification model: ResNet-IBN; Vision Transformer (DINO) with convolution stem
Dataset: two cameras w/. overlap captured around intersections of Columbia Engineering School


## Education
### Master of Science - MS in Electrical and Computer Engineering
Columbia University
Jan 2021 – Jan 2023

### Exchange Student in Computer Science
Technical University of Munich
Jan 2019 – Jan 2020


## Contact & Social
- LinkedIn: https://linkedin.com/in/humingzhe

---
Source: https://flows.cv/mingzhe
JSON Resume: https://flows.cv/mingzhe/resume.json
Last updated: 2026-03-22