I am an engineering student specializing in Data Engineering at the Faculty of Sciences of Sfax, Tunisia. Passionate about data infrastructure and distributed systems, I focus on building efficient pipelines and platforms that transform raw, unstructured data into valuable, actionable insights.
I thrive on solving complex architectural problems — from orchestrating multi-node Hadoop/Spark clusters to designing medallion lakehouse architectures. I continuously explore new methodologies in real-time processing, federated learning, and cloud-native data warehousing.
location -> Sfax, Tunisia
education -> Engineering Cycle, Data Engineering — Faculty of Sciences of Sfax
focus -> Big Data | Distributed Systems | Data Pipelines | MLOps
currently -> Building scalable data infrastructure
languages -> Arabic (Native) | English (Professional) | French (Limited)
contact -> loueylahwel@gmail.com
Programming
Big Data & Distributed Systems
Infrastructure & Orchestration
Databases
Visualization & BI
Cloud & DevOps
Automated Multi-Node Distributed Cluster Orchestrator
A self-service platform for provisioning Apache Hadoop and Spark clusters via an async pipeline. Built with FastAPI, Celery, Redis, Terraform, and Ansible. Features JWT-based RBAC, real-time status streaming, and a fully containerized deployment stack with a monitoring layer.
Stack: FastAPI Celery Redis Terraform Ansible Docker Compose
GitHub Archive Trend & Virality Analytics Platform
End-to-end analytics platform ingesting GH Archive data through a Bronze-Silver-Gold Medallion architecture on Apache Iceberg and Spark. Includes a virality scoring engine, tech-stack trend analysis, and time-travel queries backed by LocalStack S3 and Iceberg REST Catalog.
Stack: Apache Spark Apache Iceberg LocalStack S3 Medallion Architecture
Intelligent Schema-Aware Web Scraping Framework
A Python-based framework for automated schema discovery and structural HTML analysis. Features modular parsing and extraction pipelines with reusable data models for downstream analytics and warehousing — designed for production-ready data ingestion.
Stack: Python HTML Parsing ETL Pipelines Data Modeling
Text-to-SQL Local Agent (ClickHouse + LLM)
A fully local Text-to-SQL system integrating FastAPI, ClickHouse, and an LLM runtime. Implements schema introspection, SQL validation, and Dockerized multi-service deployment for natural language querying over analytical databases.
Stack: FastAPI ClickHouse LLM Docker
Federated Learning Anomaly Detection System
LSTM autoencoder models for time-series anomaly detection within a federated learning architecture. Includes preprocessing pipelines, feature engineering, and integration into distributed model aggregation workflows for network security analytics.
Stack: LSTM Federated Learning Time-Series Feature Engineering
Real-Time Analytical Data Warehouse
Azure-hosted serverless data warehouse with automated ingestion via Azure Function Apps. Applied advanced data modeling for high-frequency financial data streams. Real-time dashboards via Grafana and Apache Superset, with OLAP queries and materialized views.
Stack: Azure Azure Function Apps Grafana Apache Superset OLAP
| Certification | Issuer |
|---|---|
| Building Customized LLMs with OpenAI | Columbia+ |
| Learning AI Through Visualization | Columbia+ |
| CCNA: Introduction to Networks | Cisco Networking Academy |
| Building Data Pipelines with Apache Airflow | 365 Data Science |
| Advanced SQL for Data Engineering | 365 Data Science |
"The goal is to turn data into information, and information into insight."