About Me
Hi, my name is Nam Nguyen and I am a junior data engineer. I also have experience in developing machine learning and deep learning applications. I have worked at a number of companies gaining teamwork and leadership skills.
My technical skills are:
- Python
- Github
- Linux
My English:
- Toeic: 800
Experience
Established in Jan 2017 with mission "Changing people's lives by providing great experiences on the Internet". Teko is investing strongly in the fields of Fintech, New Retail, Logistics, Warehouse, Digital Entertainment, B2B Management Solution with greatest success is VNPAY
I have archieved experiences not only as Data Engineer but also DataOps Engineer. I have been using Python, PySpark, and open-source object storage systems to contribute:
- Providing ETL processes on big data of VNSHOP, PHONGVU, VNPAY, VINMART, POS365
- Providing reports using internal email system and Google Data Studio
- Ingesting data from multiple sources such us sequential databases, non-sequential databases, APIs, website, GCP
- Providing CI/CD using GitLab CI/CD and Apache Airflow
- Providing Avro-Parquet schema converter from json-based designs
- Providing simple automation ETL from json-based designs
Aimesoft established in 2018 with vision to providing AI products
Worked as a Machine Learning engineer, I provided scalable, reuseable computer vision applications:
- Money classification - Japan Cash Machine Co., Ltd.:
- The objective of the project is to detect realtime real-counterfeit banknotes in cash machines
- Analyzing and comparing classification algorithms and feature extractors using RGB and IR cameras
- Providing APIs for GUI application
- I used C++/OpenCV and Visual Studio 2015 in this project
- Density Analysis:
- The objective of project is detect and visualize density of people in static scenes in supermarkets, stations …etc, for better advertising
- I proposed two approaches to solve this problems:
- Person detection pretrained YOLOv3 on MOT17DET dataset. This approach provides an online, realtime solution with better accuracy
- Pre-trained CSRNET to understand crowded scenes. This approach requires knowledge about the scenes before executing, if not it would perform poorly
- I used Pytorch and OpenCV-Python in this project
Finsify is a Fintech startup from since 2014
I contributed to the company with these projects by using PyTorch, Pandas, Scikit-learn:
- Transaction classification for MoneyLover:
- The objective is to classify transaction based on their description provided by banks
- I implemented a CNN and customized Word2Vec model to do classification task. I also provide APIs for Money Lover web and application
- User segmentation for crawling optimization:
- The objective is to optimize the cost of crawling data by segmenting users in the system.
- I used KNN model with statistic features
Projects
Online Pedestrian Tracking
HUST - It is my graduation research project in which I got an A, I have learnt about doing experiments and writing scientific reports and thesis in this project
The objective of this project is tracking pedestrians while remaining their identifications during the time they appeared in the image sequences as long as possible. I proposed two solutions based on DeepSORT architecture:
- Firstly, I replaced RCNN person detector by Yolo V3 pre-trained on ImageNet.
- Second, I altered the DeepSORT architecture by using parameter-free spatial attention network as appearance descriptor.
Both proposals provide higher accuracy results comparing to original methods. But only YoloV3 proposal satisfied real-time constraint.
Error detection on cable products
HUST - A project let me get familiar with image processing in C++
We used several image processing algorithms such as binary image, Hough transform, edge detections algorithms, connected components to detect abnormal angle of connecting parts of cables. Our application can work real time with industrial camera in speed of 8 FPS.
Education
Hanoi University of Science and Technology (HUST)
Engineer's Degree of Information Technology
2014 - 2019
Founded 1956, is the first and largest technical university in Vietnam.
During my time at the university, I learnt most of my key skills that I have taken through my career such as teamwork and working to tight deadlines. I thoroughly enjoyed my time at university and learnt a lot about a healthy work life balance.
A Little More About Me
Alongside my interests in data and software engineering some of my other interests and hobbies are:
- Running
- Reading
- Pool Billiard
