Profile picture
Hello! I'm

Xuan Son NGUYEN

Engineer at Hugging Face
Graduated from INSA Centre Val de Loire

I am passionate about machine learning, low-level intricacies and all other things about computer.

About me

  • View my CV: Click here
  • Nationality: Vietnamese
  • Live in Île-de-France 🏡
  • Languages: Vietnamese - French - English
  • ✉️ Email: contact@ngxson.com
  • 📝 Biography: View here
  • Hobbies: Guitar, Painting, Photography, Graphic designing

Education

2019 - 2023

Engineer's degree (Master)

INSA Centre Val de Loire

  • Security and IT Technologies Department (STI)
  • Apprenticeship program (3 years)
  • Awarded: Valedictorian
  • Secnumedu training certified by ANSSI

2018 - 2019

BAC+1

Aix-Marseille University

  • Courses: Computer Science, Mechanics, Physics
  • Participated in Math-Info Licence 2 courses as a free auditor

2016 - 2017

Bachelor of Computer Science

Vietnam National University

  • Basic programming (in C, C++ and Java)
  • Courses: Mathematics and Physics

What can I do?

  • C, C++, Python : system programming, machine learning
  • Javascript and Typescript
  • Java, Kotlin (for Android)
  • Assembly: reverse engineering
  • Docker and Linux security (AppArmor, SELinux,...)

Other things:

  • Browser extensions development (Chrome / Firefox)
  • Facebook Messenger bot development
  • Answer to questions on StackOverflow

Working experiences

Software engineer at Hugging Face
August 2024 - now

Software engineer at Hugging Face

Hugging Face is a machine learning and data science platform and community that helps users build, deploy and train machine learning models.

My works:

  • Working with On-device LLM and runtime (llama.cpp, ONNX)
  • Maintaining the integration of GGUF / llama.cpp on the Hugging Face hub

Website: huggingface.co

Network and system engineer at Snowpack: C++ and Flutter developer
September 2022 - August 2024

Network and system engineer at Snowpack: C++ and Flutter developer

Snowpack is a startup that combines privacy and security on the internet using an overlay network to ensure users are truly invisible online.

My works:

  • Participating in the development of Snowpack in C++ and Flutter.
  • Optimizing of system performance and stability. Bug / vulnerability fixes.
  • Participating in the certification process of the Snowpack solution, development of unit tests and applied documentation.
  • Engineering and maintaining the platform API server

Skills : C, C++, Python, Typescript (API server), Flutter (desktop app), Swift (iOS), Kotlin (Android), Network Security, Docker

Website: snowpack.eu

Apprentice at Botfuel: Full-stack web developer
September 2020 - September 2022

Apprentice at Botfuel: Full-stack web developer

Botfuel is a startup specialized in conversational intelligence. We offer virtual assistance solutions based on artificial intelligence to optimize the customer relationship management of brands. Our virtual assistance can be installed on a brand's existing e-commerce site.

My works:

  • Developing virtual assistance solutions based on artificial intelligence.
  • Building and maintaining the chatbox (chat window) widget that can be installed on e-commerce sites of our clients.
  • Implementing integrations with external channels including Facebook Messenger and Zendesk.

Skills: Flask (Python), Docker, ReactJS, websocket

Website: botfuel.io

Internship at SimpleLogin: Browser extension developer
July 2020

Internship at SimpleLogin: Browser extension developer

Open-source software

SimpleLogin is an open-source service to help you protect your email address by giving you a permanent alias email address. SimpleLogin is currently a subsidiary of ProtonMail.

My works:

  • Developing the extension that allows users to generate an alias email address and fill out forms automatically.
  • Optimizing the user experience of the extension.

Skills: Browser extension development (Chrome / Firefox), VueJS

Website: simplelogin.io

Personal projects

Collaborator of llama.cpp

2024 - now
llama.cpp enables LLM inference with minimal setup and state-of-the-art performance on a wide variety of hardware.
Project website: https://github.com/ggerganov/llama.cpp

Stack-based virtual machine visualization

Feburary 2022
For using in "Compilation" course at INSA Centre Val de Loire. This project has been used by my teachers since 2022. It provides a user interface for teachers and students to learn how to write assembly code for a stack-based machine.
Live demo: https://edu-insa-4a-compilation-vm.netlify.app/

Amazon Kindle Jailbreaking

2021 - 2022
This project aims to unleash the power of the Amazon Kindle e-book reader. This is done by “jailbreaking”, or taking the root access of the device.

Facebook Messenger bot (inactive)

2017 - 2020
This Messenger bot allows high school students to meet and chat with other high school students they don’t know.

  • Used by: around 10,000 users
  • Technologies used: NodeJS, MongoDB, Firebase and AngularJS

Android Application "Ban co biet" (Collaboration)

2016 - 2018
Application offering verified and fun anecdotes on various subjects: science, history, cinema, sport, etc.

Custom Linux kernel for Sony Xperia M (inactive)

2014 - 2017
Kernel allowing users to access advanced features of their Sony Xperia M smartphone (overclocking, I/O control, etc.)

Latest news

2024: LLM - Hugging Face - New Horizons

As 2024 began, I embarked on a new chapter of self-discovery and growth, dedicating my free time outside of my role at Snowpack to contributing to open-source projects as a means of continuous learning. One project that particularly captured my interest was llama.cpp, an innovative initiative enabling the deployment of large language models locally, without relying on external servers.

As I began contributing to llama.cpp, I was thrilled to see the project grow and gain momentum. My efforts did not go unnoticed, and I was approached by Julien Chaumond, CTO of Hugging Face, who expressed interest in collaborating with me on on-device LLM technologies such as ggml, ONNX, and more.

I was honored to accept the offer and joined Hugging Face in August 2024. I was eager to leverage this opportunity to expand my knowledge in LLM and AI, and I was not disappointed. Working alongside a talented team, I was immersed in a dynamic and innovative environment that fostered growth and exploration.

One of the highlights of my tenure at Hugging Face was the launch of the Ollama compatibility layer in October 2024. In collaboration with my colleagues, we successfully bridged the gap between two prominent LLM platforms, enabling users to utilize GGUF models from Hugging Face directly on Ollama. In the 24 hours following the release, we received over 20,000 requests from users to download a wide variety of models. This achievement was a testament to the power of collaboration and innovation, and I was proud to have played a part in it.

Photos: (left) me in front of Hugging Face office in Paris, (right) me at Entrepreneur First's hackathon event. Thanks to Marc Sun and Kevin Kuipers for taking these photos 🤗

As I reflect on my time at Hugging Face thus far, I am filled with a sense of accomplishment and gratitude. I am excited to continue contributing to the domain of LLM and AI, and I look forward to the opportunities and challenges that lie ahead.

[Read more on my biography]

This portfolio is coded by Xuan Son NGUYEN, using NextJS and TailwindCSS

Printable version: Click here