Recent engineering graduate in Knowledge Engineering and Data Science, passionate about building data pipelines and full-stack, AI-powered products. Hands-on experience across ETL/ELT, data warehousing, visualization, and NLP, as well as full-stack development and AI integration (LLMs, RAG, agents). Immediately available for new opportunities.

ESI curriculum: Artificial Intelligence · Machine Learning · Deep Learning · Big Data · Cloud Computing · DevOps · Software Engineering

• Built and maintained an on-premise, multi-tenant data factory using Meltano, dbt, and Podman
• Ingested 807+ tables and 58M+ rows (20 GB), reducing processing time by 89%
• Translated functional specifications into technical data models, working with architects and business stakeholders in Agile delivery

• Developed full-stack features for external clients: Next.js frontend and backend logic for AI-powered services
• Integrated an AI chatbot (LangChain, Qdrant, OpenAI API) and a collaborative-filtering recommendation engine
• Designed backend RAG pipeline logic and deployed services on AWS ECS with CI/CD via GitHub Actions

• Built a pipeline extracting and consolidating data from 5 internal systems (14 GB)
• Designed a self-serve dashboard adopted by 9 stakeholders, saving 2 hours per day

• Cleaned, transformed, and analyzed large-scale social media data
• Contributed to fine-tuning NLP models for localized social listening in Morocco

Intelligent chatbot integrated with WhatsApp, capable of querying and managing multimedia data (text, image, audio, video); validated with 51 test users

• Multi-agent architecture with LangChain / LangGraph for task distribution, orchestration, and contextual persistence
• RAG system powered by Qdrant for semantic search, vector indexing, and conversational grounding
• Scalable and secure data pipelines on AWS (S3, DynamoDB, Lambda) — ingestion, storage, transformation, monitoring
• Multi-turn conversational context with OpenAI APIs for NLU and data synthesis

Stack: Django REST · AWS · Qdrant · LangChain · OpenAI API · GitHub Actions

End-to-end data pipeline based on the Medallion architecture

• Extracted, processed, and analyzed data from clients' on-premises databases (~900 tables, 20 GB per client)
• Leveraged Microsoft Fabric's unified platform — Lakehouse, OneLake, Data Pipelines, Dataflow Gen2 — to structure Bronze, Silver, and Gold layers
• Configured JDBC-based connectivity for secure ingestion, refreshing 3x/day (40 min full load, 5 min incremental)

Stack: Microsoft Fabric · Lakehouse · OneLake · JDBC

End-to-end ETL pipeline into a star-schema OLAP warehouse

• Moved ~50 GB of NYC Green Taxi trip data into dimension and fact tables optimized for analytics
• Configured an Apache Airflow DAG with monthly scheduling, automated data validation, and automatic schema creation
• Optimized for scale with chunked inserts, connection pooling, and indexing; automated backfill across 71+ months of historical data, with full pipeline runs completing in ~5 minutes

Stack: Apache Airflow · Spark · Python · SQL

Source — Databricks migration of the on-premises NYC Taxi pipeline into governed lakehouse data products for fleet operations, finance, and compliance

• Rebuilt the pipeline on Databricks + Unity Catalog, using Auto Loader for incremental ingestion into a Bronze/Silver/Gold/Quarantine medallion architecture on Delta Lake
• Implemented data-quality rules (invalid timestamps, zero-distance trips, duplicate business keys, negative fares) with quarantine tables for auditability instead of silent data loss
• Built Gold dimensional models (zone, date) and daily borough performance metrics via Databricks SQL Warehouses
• Deployed and orchestrated jobs with Databricks Asset Bundles, replacing the original Docker/Terraform-based on-premises setup
• Published 3 AI/BI dashboards (Fleet Operations, Financial Performance, Compliance & Quality) consumed by distinct business stakeholders

Stack: Databricks · Unity Catalog · Delta Lake · Auto Loader · Databricks SQL · Databricks Asset Bundles · PySpark

End-to-end streaming data platform for live market analytics

• Ingested live BTC-USD prices into Apache Kafka, processed with Spark Structured Streaming for per-batch model training
• Developed a FastAPI backend (API + interactive Streamlit dashboard) exposing real-time model metrics and batch predictions
• Orchestrated a multi-container architecture (Kafka, Zookeeper, Spark, dashboard, API) with Docker Compose, with monitoring and metrics persistence

Stack: Kafka · Spark · FastAPI · Streamlit · Docker Compose

Enterprise data warehouse and analytics platform for banking operations

• Designed and implemented a star schema Data Warehouse on SQL Server
• Built ETL processes with SSIS for ingestion and transformation of complex banking data
• Created interactive Power BI dashboards with business KPIs and predictive insights
• Optimized DirectQuery / Import connections between Power BI and SQL Server for real-time analytics

Stack: SQL Server · SSIS · Power BI · Python

Full mathematical implementation of ML and DL algorithms

• 6 ML models: Linear Regression, Logistic Regression, KNN, Decision Trees, Random Forest, SVM
• 4 DL architectures: MLP, CNN, RNN, Autoencoder
• Complete backpropagation and optimization algorithms implemented from scratch
• Evaluated on real datasets with comprehensive performance metrics

• Co-Head, Competitive Programming Cell — CODE-ESI (Sep 2024 – Jun 2025) Mentoring engineers in algorithmic problem-solving

• Treasurer — JCMP-ESI (Sep – Dec 2024) Managed club finances and budgeting

• Sponsorship & Events Committee Member — Moroccan Days of Future Data Scientists (May 2024 – Present) Coordinated partnerships for national data science events