Profile
Data Engineer with 2+ years building humanitarian data infrastructure at IMPACT Initiatives — Python ELT pipelines, PostgreSQL star schemas, and Power BI dashboards used by field teams in Sudan and Libya. MSc in Data Analytics (16.5 / 20). Finance background and 3 years teaching statistics give a strong analytical foundation. Ready to own the data stack end-to-end — from ingestion to insight — in a team where data quality directly drives decisions.
Experience
Data Engineer — Sudan Mission
Jun 2024 – Present
IMPACT Initiatives · Nairobi, Kenya
- Designed and built a PostgreSQL star schema consolidating 10+ humanitarian data sources
- Developed Python ELT pipelines ingesting field data from KoboToolbox API, with spatial joins, schema validation, and automated transformation
- Built a 4-tier location matching system achieving 100% geographic coverage across Sudan
- Created 10+ fact & dimension views supporting 26+ months of time-series field data
- Delivered Power BI dashboard used by humanitarian teams to track displacement and service gaps across 55 admin clusters
Data Officer — Libya Mission
Feb – Jun 2025
IMPACT Initiatives · Tunis, Tunisia
- Processed and cleaned 5,000+ survey responses via KoboToolbox across 3 assessment rounds
- Built Power BI dashboards tracking displacement, food security, and healthcare access across 8 governorates
- Ran vulnerability and trend analysis in R and Python, outputs feeding directly into IMPACT situation reports
- Trained 5+ partner organisations on KoboToolbox and data quality best practices
Data Scientist Intern
Feb – Jun 2024
VERMEG for Banking & Insurance Software · Tunis, Tunisia
- Automated email classification using RoBERTa and DeBERTa achieving 92% accuracy
- Built BART summarisation pipeline cutting manual email review time for compliance teams
- Built sentiment classifier to triage inbound tickets by urgency, reducing time-to-first-response on critical issues
BI Analyst Intern
Jul – Aug 2023
Wimbee · Tunis, Tunisia
- Built Power BI dashboard giving the sales team real-time pipeline visibility, replacing manual weekly reports (15% efficiency gain)
- Rebuilt Talend ETL pipeline integrating CRM and transaction data into the reporting layer, eliminating manual data exports
Projects
Sudan Humanitarian Data Infrastructure
↗ IMPACT Initiatives
Field assessment data across Sudan was siloed across 10+ disconnected sources with no geographic standardisation. Built a Python ELT pipeline (KoboToolbox API → PostgreSQL star schema) achieving 100% location coverage, with Azure Blob archiving and a Power BI dashboard serving humanitarian teams across 55 admin clusters.
Smart Box — AI Email Management System
↗ GitHub
Email triage at fintech firms creates compliance bottlenecks when handled manually. Built a 5-model NLP pipeline automating classification (RoBERTa), sentiment scoring (BERT), NER, summarisation (BART), and response generation (LLaMA-2) — containerised with Docker and deployed on HuggingFace Spaces.
Olist Sales Analysis Award Winner
↗ GitHub
With 99K+ unstructured transactions, identifying at-risk customers was guesswork. Designed a full analytics stack — Talend ETL into a star schema warehouse, Power BI reporting, and an XGBoost model predicting customer satisfaction at 88.4% accuracy. Award-winning at ESPRIT Bal de Projet 2023.
National Health Analysis
↗ GitHub
Hospital management had no unified view of patient flow and costs across 10 facilities. Built a Power BI dashboard across 55,500 records surfacing admission trends, billing patterns, and treatment outcomes to support resource planning.
Education
MSc Data Analytics
ESB – Esprit School of Business
2022 – 2024 · 16.5 / 20
BSc Business Management (Finance)
MSB – Mediterranean School of Business
2018 – 2021 · GPA 3.16 / 4
Also served as Statistics & Mathematics Teaching Assistant · MSB · 2018–2021