Head of Data at WalaPlus. Data platforms on Google Cloud, data governance, applied ML.
Also a part-time AI Researcher at Momita.ai.
Before that: Data Engineering and Machine Learning Lead at Rowad Modern Engineering, and a first career in civil-engineering project controls, which is where the data work started.
omaressam.web.app · LinkedIn · Credly · Kaggle · Hugging Face
| Repository | What it shows |
|---|---|
| telco-churn-classification | Six classifiers compared on the same folds; logistic regression chosen for interpretability (AUC 0.843); odds ratios and a gains chart for a retention campaign |
| customer-segmentation-kmeans | Feature engineering doubles the silhouette; five segments, one of them holding 65% of churned revenue |
| fashion-mnist-cnn-gradcam | MLP to CNN (93.7%), a confidence gate that auto-tags 80% of images at 99.1%, Grad-CAM |
| course-recommender-system | Content-based, clustering and collaborative-filtering recommenders on 233k enrolments, plus three bug fixes in the reference models |
| backend-nodejs-capstone | Node.js / Express / MongoDB marketplace API with JWT auth, a sentiment microservice, Docker Compose and CI |
| hieroglyph-classifier | A CNN that names Egyptian hieroglyphs by Gardiner code: 96.9% top-1 on 59 signs from the Pyramid Texts of Unas; model and demo on Hugging Face |
| pharaohs-of-egypt | A reproducible pipeline from Wikidata to validated tables of pharaohs, dynasties and pyramids, published as Kaggle and Hugging Face datasets |
| databricks.project · data.warehousing.capstone · snowflake.tasty · database.admin.capstone | Data engineering: a Databricks medallion lakehouse with Structured Streaming and Unity Catalog, an OLTP-to-star-schema warehouse in MySQL and PostgreSQL, Snowflake ingestion and Snowpark, Airflow ETL |
| nile-river-flow | Change-point analysis of the Nile at Aswan, 1871–1970: three methods agree on 1898–1900, and an honest test of what a forecaster gains |
| superapp | ARIA, an agent framework on LangGraph with a FastAPI back end and a React Native client |
| graph-rag | Graph RAG: spaCy entities into a NetworkX knowledge graph, retrieval from the graph to ground LLM answers |
| aavail-revenue-forecasting | IBM AI Workflow capstone: 30-day revenue forecasts, a fair backtest of ten methods against a naive baseline, a Flask API in Docker, 24 tests in CI |
| django-online-course · rag-qa-bot-langchain | A Django app with an exam feature and tests; a LangChain RAG bot over PDFs with local models |
| spacex-falcon9-landing-prediction | IBM Data Science capstone: SpaceX API and Wikipedia scraping, SQL and visual EDA, a Folium map, a Plotly Dash dashboard and four tuned classifiers (83% on held-out launches) |
| data-science-with-r · data-analytics-capstone-cyclistic | Analytics projects in R and Python |
| e-plant-shopping · travel-recommendation · express-book-reviews | Front-end and API course projects: a React/Redux shop and a vanilla-JS site, both live on GitHub Pages, and an Express REST API with JWT |
Write-ups of the portfolio projects, on dev.to (also on Medium and Hashnode):
- Teaching a CNN to read hieroglyphs: 96.9% on 59 signs, and why my first model failed
- Pharaohs of Egypt: a validated dataset from Wikidata, built like a data pipeline
- The year the Nile changed: change-point detection on 100 years of river flow
- Churn prediction where the simple model wins: logistic regression against five rivals
- Five customer segments, and the one that holds 65% of the lost revenue
- An image classifier you can trust 80% of the time: confidence gating and Grad-CAM
Platform: Google Cloud (BigQuery, Cloud SQL, Cloud Run, Cloud Build, Pub/Sub, Dataflow, Vertex AI), Docker, Terraform, GitHub Actions Data: PostgreSQL, MySQL, BigQuery, Snowflake, Redis, MongoDB; Airflow, dbt, Spark, Kafka; dimensional modelling, data quality and governance ML: scikit-learn, XGBoost, PyTorch, TensorFlow, RAG and agent frameworks (LangChain, LangGraph) Languages: Python, SQL, JavaScript/TypeScript, R
352 Coursera certificates (18 professional certificates, 47 specializations) and 134 Credly badges, every one with a public verification link on the credentials page. Currently preparing for the Google Cloud Professional Data Engineer, Cloud Architect and Cloud DevOps Engineer certifications.