Amine Harrabi

I’m a data engineering student building a strong foundation in data structures, databases, machine learning, statistical analysis, and computer networks, and I enjoy turning messy real-world inputs into reliable data products. I care about systems that are maintainable, measurable, and designed for continuous improvement.

In my Data Engineering internship, I designed a production scraping pipeline that automatically collects textures, 3D models, PDF schemas, and color data to keep a core product up to date. I also migrated an 8M+ row PostgreSQL database to Elasticsearch using custom index mappings, cutting search latency from around 15 seconds to near-instant. On top of that, I built and documented RESTful Elasticsearch API endpoints and developed scalable Python data-processing workflows to support ongoing product maintenance.

Outside the internship, I keep expanding my engineering range through hands-on projects: a real-time C++/OpenGL plasma simulation with compute shaders and interactive controls, and an image generation pipeline (RQ-VAE + UNet) trained with perceptual loss and mixed precision. I’ve also worked on practical data workflows like dialogue dataset sanitization and merging for LLM fine-tuning (including LoRA with 4-bit quantization), plus tooling that wraps and routes external APIs into clean REST services. Overall, I’m motivated by building end-to-end pipelines—from ingestion and transformation to serving and performance—while learning how to apply the right tools for each problem.