About me

A backend engineer with 3 YoE in software engineer field. I am goal-oriented, dedicated to my work, and always open to new experiences, especially those related to my field of study, to enhance my coding practice.

I have gained many from being a leader of group projects, which is why problem-solving, presentation and leadership skills are not weaknesses of mine. I have a strong passion for software engineering and computer science, and I hope I can contribute to the development of society.

Education

  • Bachelor in Information Technology
    2019 - 2023

    Major in Artificial Intelligent (Computer Science)
    Valedictorian graduated (Top 1 with GPA 3.94/4)

Working experience

  • Software Engineer April 2023 - Now

    Design and develop large-scale zero-downtime systems with millions users. Using Java Spring & distributed systems on Naver cloud to implement social media platforms (Naver BBoom and Naver Blog). Responsible with DevOps task (e.g. build CI/CD pipeline). Building Slack chatbot with Slack Bolt to monitor the service when deploying.

  • Freelance Software/AI Engineer Nov 2022 - March 2023

    Develop software tools and utilities (e.g. syntax checker, trading probabilistic calculator, simple compilers and algorithms to solve large data problems). Build dashboard to visualize data, analyze and provide insight from the dashboard. Build machine learning models and fine-tune them to achieve a good measure metrics (accuracy, F1-score, etc)

  • Fullstack Blockchain Intern May 2022 - Oct 2022

    Research blockchain technologies and develop a blockchain-based application to digitalize real estates. Focus mainly in backend development and smart contract implement with NEAR and Aurora Engine.

  • Programming Lecturer Nov 2020 - Apr 2022

    Teach programming & STEAM for children in the age of 12 to 18 year old. Tutoring in algorithms using Python language & application development with frameworks (e.g. Python Turtle, PyQt5, Wordpress).

  • Mobile/Backend Developer 2019 - 2020
    CodeNGreen Ltd | Vietnam market

    Develop mobile application and backend servers which connect with IoT device to automatically water gradens.

  • Embedded Software Researcher Apr 2018 - June 2018

    Research embedded systems and write instruction blogs about designing softwares with IoT devices. Setup machines for AI labs (e.g. Raspberry Pi, 3D Printer).

Skills

Project

naver-blog
Blog (블로그)
Software | Naver

Naver Blog is a social network which used to upload blogs. The service is used by many people in Korea, which has an extremely high traffic. In this project, we developed the map feature, which display blogs on Google Map & Naver Map. Moreover, we improved the RSS system.

naver-bboom
BBoom (뿜)
Software | Naver

A social network which used to upload humour posts and memes, titled as Naver BBoom. With millions of user, we are maintaining and improving all it's services. A lot of legacy features were optimized by us which reduce the latency and make the service more secure.

huimitu
Huimitu
Software | University

Huimitu, an e-commerce website which integrated with payment services (e.g. Momo, Paypal). Furthermore, machine learning has been applied to do product recommendation. There are many other interesting feature in the project (e.g. geocoding reverse, statistic visualize).

dqn_snake
DQN snake
Machine learning

Implement a snake game as the training environment for Deep Q-learning Network (DQN) model. The DQN then integrated with A* algorithm to help my agent solve the game. After training & fine-tune, my agent achieved the 1st prize in the university contest.

behaviour
Behaviour classification
Machine learning

An AI model which can detect user's pose from real-time camera then classify user's action. Compared to traditional action classification, this model can use frame-speed to distinguish similar actions (e.g. walk vs run) by leveraging MediaPipePose and LSTM.

rcnn
RCNN for car
Machine learning

We researched about the RCNN model, which can detect multiple objects in an image (instead of a single). By reading papers, we implemented the model and deploy it using StreamLit to do the inference through the website GUI.

Awards & certificates

Anually received scholarship excellent students. Graduated valendictorian with the highest GPA (3.94/4) in the Information technology department over all the programmes.

Top Rated highlights professionals who have built a strong reputation on Upwork by getting positive feedback from their clients, time after time. They represent the top 10% of talent on Upwork.

British Council

IELTS is an English language proficiency exam that reports at all levels from low intermediate to very advanced. There is a general and academic version.

GFS Blockchain

A course on how to use NEAR Protocol to write smart contracts and integrate to an application. Furthermore, Aurora Engine and Octopus Network are also covered in the course. At the end of this course, learners must propose an software that use blockchain and build the prototype.

STEP STEM Alive 2019 - 4th runner-up

The programme sought to promote interest in STEM education and harness creativity in translating cross-curriculum knowledge into solutions for real-world issues to prepare participants to be self-directed learners, active solution contributors to the United Nations Sustainable Development Goals.

Earners of this certificate have passed the L1 examination of KNIME Certification Program. Thereby they have demonstrated knowledge of the following topics: General concepts of KNIME Analytics Platform, importing/exporting data, data manipulation, and data aggregation.

Shopee Code League is a 2-week coding league consisting of two rounds of coding competitions, open to all students and professionals. Participants will put their coding skills to the test by solving challenging algorithmic puzzles designed by the Shopee Engineering team.

Publications

Propose a method to improve current action spotting vision models by unifying global and local feature. The method not only help AI model improving the accuracy but also more explainable By applying the GLIP - a zero-shot learning model of Google, the method can apply on arbitrary sports.

Story about how I built a recommendation system for an e-commerce (named Huimitu) system, using KNIME.

- Do Vuong Phuc © 2024 -