Ghassen Boujlida
Machine Learning Engineer
Machine Learning Engineer specializing in recommendation systems, personalization engines, and scalable ML infrastructure. Also building modern web experiences.
01. Experience
Where I've Worked
Machine Learning Engineer @ SODA inc. (SNKRDUNK)
Dec 2025 - Present
- Joined as the first Machine Learning Engineer, building the ML infrastructure from the ground up
- Developing recommendation and personalization systems for the sneaker marketplace platform
- Built an LLM agent using LangGraph to extract business-relevant keywords from product names, powering the recommendation section of the app
Machine Learning Engineer @ LY Corporation (LINE)
Apr 2023 - Nov 2025
- Implemented recommendation logic for LINE app official account suggestions, achieving 110% increase in CVR and +200M yen yearly revenue
- Built user-interest vector embeddings (dense & sparse) using PySpark, Python, and PyTorch, used across multiple LY Corp products
- Developed a high-performance Go-based API for personalized search result reranking in LINE Official Accounts
- Conducted PoC using Vision LLMs for image feature extraction, fine-tuning models in Japanese
- Optimized models by merging separate ML tasks into a single model, reducing maintenance costs
- Enhanced internal ML library, boosting task accuracy by up to 6%
- Implemented automated data analysis using LLMs with exceptional results
Research Intern @ NEC Corporation
Aug 2021 - Oct 2021
- Contributed to Kurumie for Cities road damage detection service
- Trained models from scratch and implemented a new technique that enhanced detection accuracy by 7%
Software Engineer (Part-time) @ Rutilea, Inc.
Dec 2020 - Mar 2023
- Developed an industrial bolt detection service using computer vision, successfully deployed to production
- Created PoCs using Neural Radiance Fields (NeRF) for 3D modeling, doubling the modeling speed
02. Projects
Things I've Built
Featured Project
Japan Easy Travel
A full-featured travel company website for a Japan-based boutique travel agency. Features include tour listings with filtering, destination pages, private guide booking, testimonials, and a fully responsive design optimized for conversion.
LINE Recommendation System
Built the recommendation logic for LINE app official account suggestions, resulting in a 110% increase in conversion rates (CVR) and an additional 200 million yen in yearly revenue.
User Vector Embeddings Platform
Designed and maintained user-interest vector features (dense & sparse representations) capturing diverse behavioral signals, now used across multiple LY Corp products for personalization.
03. Skills
Technologies I Work With
Languages
- Python
- Go
- SQL
- C++
- TypeScript
ML / Data
- PyTorch
- PySpark
- TensorFlow
- LLMs
- LangGraph
- Computer Vision
- NLP
Infrastructure
- Docker
- Kubernetes
- Argo Workflows
- ArgoCD
- GitHub Actions
- Grafana
- AWS
- Terraform
Web
- Next.js
- React
- Tailwind CSS
- Node.js
04. About
Background
A Machine Learning Engineer based in Japan since 2015, when a MEXT scholarship from the Japanese government opened the door to a career in tech across cultures and languages.
Master's degree in Statistics and Machine Learning from Osaka University, with research focused on density estimation with autoencoders and normalizing flows. Professional experience spans research at NEC, production ML systems at LINE (LY Corporation) serving millions of users, and currently the first ML engineer at SNKRDUNK (SODA inc.).
Beyond ML, experienced in full-stack web development — bridging data science and end-user experiences. Fluent in four languages and comfortable working across multicultural teams.
Education
M.Eng in Statistics & ML, Osaka University
B.Eng in Mathematics · KOSEN Nagano
Languages
Scholarships
MEXT Scholarship (6 years)
Rotary Yoneyama Memorial
05. Contact
Let's Work Together
Open to new opportunities, collaborations, and interesting projects. Whether it's an ML system, a web application, or a conversation — don't hesitate to reach out.