Yahya Hadi
I’m Yahya Hadi, a Data Science and AI engineer trained at EPFL, where I built end-to-end machine learning and data pipelines and turned complex technical ideas into useful, user-friendly systems.
About
I’m Yahya Hadi, a Data Science and AI engineer trained at EPFL, where I built end-to-end machine learning and data pipelines and turned complex technical ideas into useful, user-friendly systems. My experience spans NLP, data engineering, and deep learning, including building classification and retrieval pipelines, designing ML workflows with modern tooling, and optimizing models through experimentation and cross-validation.
I enjoy bridging research and production by focusing on robust data lifecycles, scalable systems, and clear results. From wrangling historical datasets and building interactive visualizations, to developing synthetic data generation and real-world time-series classification for industrial applications, I aim to deliver accurate models that are practical to deploy—supported by strong MLOps/DevOps practices.
- 5 years' experience
- Joined July 2026
Skills· 15
Languages
- FrenchFluent
- ArabicFluent
- EnglishFluent
- Spanish; CastilianBeginner
Industry experience
- Consumer Goods
- Computers & Electronics
- Education
- Professional Services
- Software & Internet
- Media & Entertainment
More experts in Lausanne, Switzerland
Work experience· 6 roles
Built a synthetic data generation pipeline using the Blender API to create diverse datasets for gaze estimation. Translated physiological phenomena into technical requirements to design a CNN architecture and inference pipeline for accurately detecting cognitive load and fatigue. Delivered modern end-to-end ML systems for time-series classification using IMU data, covering data collection, experiment design, preprocessing, and model development; optimized deep learning models under cost and memory constraints and deployed robust DevOps/MLOps workflows using AWS SageMaker and MLflow to enable continuous integration of models and edge AI readiness.
Developed a NLP framework to classify international patents into 16 UN SDGs; built an NLP pipeline to transform abstracts into model-ready features using NLP techniques and NN architectures; created Sankey diagrams to visualize correlations between CPC identifiers and predicted SDG impact.
Architected a synthetic data generation pipeline via Blender API to produce diverse datasets for gaze estimation; developed CNN architectures for gaze estimation; delivered end-to-end ML system for time-series classification using IMU data; optimized performance with CNNs and feature engineering.
Education
Hire Yahya Hadi today
Post a job and invite Yahya, or contact them directly. No hidden commission.