Welcome to My Portfolio

My name is Hesham Rashid, this is my personal portfolio website. I created this to show some of my personal projects and course work. Feel free to check out some of my projects and some of the live demos. Not a technical person? Just ask Ember, the AI portfolio assistant in a project and he will help simplify the project description and go through relevant skills.

View My LinkedIn View My GitHub Download Resume

My Projects

Check out my portfolio of projects including database systems, C# applications, and machine learning work.

View Projects

My Courses

Check out relevant courses that I've taken in both my Bachelor's and Master's program, covering data analytics, databases, and more.

View Courses

Featured Projects

A selection of my best work, click to learn more.

Recipe Vault App and MunchAI

The Recipe Database is a relational database system designed to manage and organize culinary data across a normalized 17 table schema, supporting recipes, ingredients, instructions, tags, cuisines, units, and categories. The schema enforces referential integrity through a network of primary and foreign key relationships, enabling complex multi-table queries that retrieve complete recipe profiles including ingredient quantities with units, step-by-step instructions, cuisine classifications, and user-defined tags in a single join. Designed with scalability in mind, the data model cleanly separates concerns between lookup tables such as Units, Tags, and Cuisine and transactional junction tables such as RecipeIngredients and RecipeTags, following third normal form principles to eliminate redundancy. The accompanying SQL query library covers a full range of development and analytical use cases including complete recipe views, ingredient and tag-based searches, cuisine filtering, and aggregate reporting on recipe distribution by cuisine type. Documentation includes a detailed data dictionary specifying data types, size constraints, null policies, and indexing strategies for every column across all 17 tables, alongside an entity-relationship diagram illustrating the full schema architecture. The project demonstrates relational database design, query optimization with selective JOIN strategies, schema documentation best practices, and the ability to model a real-world domain with complex many-to-many relationships.

View Details

Grocery Store Sim

The Grocery Store Simulation is a full-stack data engineering project that demonstrates a complete OLTP to OLAP pipeline through a self-running mock grocery store. The system simulates a 14-hour store day with realistic customer traffic patterns, including rush hour spikes, loyalty vs walk-in customers, and automatic inventory reordering, generating hundreds of transactions per simulated day against a normalized PostgreSQL transactional schema with 8 tables covering products, inventory, employees, customers, transactions, and purchase orders. At the end of each simulated day, a custom ETL pipeline migrates the transactional data into a star schema data warehouse, transforming raw OLTP records into a fact_sales table surrounded by dimension tables for time, product, customer, employee, and payment method, with pre-calculated revenue, cost, and profit measures optimized for analytical queries. Built with Python and PostgreSQL, the project includes a Flask-powered live transaction feed showing real-time inserts as they happen, a crash recovery system using state persistence to handle interruptions cleanly, and a suite of analytical SQL queries covering daily revenue trends, category performance, rush hour analysis, and customer spend comparisons. The simulation runs continuously and headlessly on a Raspberry Pi Zero 2 W acting as a mock server, with the OLAP database accessible remotely for reporting, demonstrating practical implementation of data warehousing concepts including ETL design, dimensional modeling, and the separation of transactional and analytical workloads across a real distributed system.

View Details

Grocery Simulation Databricks

Grocery Store Databricks is a data engineering portfolio project that demonstrates a production-style Lakehouse pipeline built on Databricks, processing 6,668 simulated days of grocery store transactional data through a three-layer Medallion Architecture. The dataset originates from a custom PostgreSQL OLAP simulation previously running on a Raspberry Pi, exported as a star schema consisting of five dimension tables and one fact table at the line-item grain, totaling over 700MB of raw data. The pipeline is implemented across three Databricks notebooks that sequentially ingest, clean, and aggregate the data using PySpark and Delta Lake, with each layer stored as managed tables in Unity Catalog under a dedicated catalog and schema hierarchy. The Bronze layer ingests raw CSVs from a Unity Catalog Volume into Delta tables using explicit schema definitions to enforce data integrity at the source. The Silver layer applies data quality filters, standardizes string formatting, derives calculated measures including profit margin percentage, and denormalizes dimension labels onto the fact table to produce a single enriched source of truth. The Gold layer produces seven business-level aggregate tables covering daily revenue and profit trends, product category performance, rush hour versus non-rush hour sales patterns, loyalty versus walk-in customer behavior, top products by profit, payment method breakdown, and employee performance ranked within role. The project demonstrates end-to-end data pipeline development with PySpark, Delta Lake table management and schema enforcement, Medallion Architecture design principles, and hands-on experience with modern Databricks features including Unity Catalog and Volumes.

View Details

Simuleras, A Simulation Website

This is a full-stack simulation web application that models the daily operations of an amusement park through a randomly generated, analytics-driven report engine, that generates a PDF analytical report, as well as CSV and SQL files of simulated data. The main goal of this project was to create a PRD (Project Requirements Document) for Claude Code to build and follow the requirements, as well as an application to generate clean data for people to do analytical reports with. The PRD can be seen in the downloads of this project. The system accepts a user-defined simulation period of up to 365 days and models each day with a dynamic weather system, realistic visitor attendance patterns, and six distinct visitor archetypes, families with young kids, thrill seekers, teenage groups, couples, seniors, and school groups, each with unique ride preferences, spending behaviors, arrival windows, and group compositions that drive variation across every simulated day. The park consists of seven themed rides with individual capacity limits, hourly throughput tracking, and random breakdown probabilities that trigger maintenance incidents logged throughout the report, alongside twelve food, beverage, and retail locations each with their own transaction values and service capacities. Revenue is tracked across three streams, ticket sales across four pricing tiers, food and beverage, and retail merchandise, with all simulation logic running entirely in memory through a Python backend with no database attached, meaning every run produces a completely fresh and unique dataset. At the end of each simulation, a fully generated analytical report is rendered in the browser covering executive summary KPIs, day-by-day attendance and revenue trend charts, ride utilization rankings, store performance breakdowns, visitor demographic distributions, weather impact analysis, and a complete incidents log, with the full report available as a downloadable PDF. Built with React and Vite on the frontend, FastAPI on the backend, and Recharts for data visualization, the project demonstrates practical implementation of simulation-based data generation, in-memory ETL aggregation, multi-stream revenue modeling, and full-stack report delivery without any persistent storage layer.

View Details

Downloadable Software

These projects have a released build you can download and run directly.

Pixel Aquarium Widget

Pixel Aquarium is an open source desktop widget that renders a living pixel art aquarium directly on the user's desktop, running entirely in a lightweight Electron shell with no cloud dependency or external services. The application features a procedurally animated ecosystem including three species of pixel art fish with individual blink cycles and swimming behavior, a pixel castle, coral formations, rocks, seaweed, and a continuous bubble system, all rendered on an HTML5 canvas using vanilla JavaScript with no runtime dependencies. A full day/night cycle shifts the sky, water color, sun, moon, and stars across an eight-minute loop, giving the widget a persistent ambient presence on the desktop. Built with a transparent, frameless Electron window configured to sit below all active application windows. Distributed as a cross-platform desktop application, the build pipeline leverages GitHub Actions to automatically compile and publish Mac and Linux installers on every release, while the Windows installer is packaged locally using electron-builder with NSIS targeting.

View Details Release

NullCloud

NullCloud is an open source desktop application that brings AI-powered coding assistance entirely to the local machine, eliminating cloud dependency and data privacy concerns by running all inference through Ollama on the user's own hardware. The application features a three-tool interface providing a SQL Query Builder, Code Creator, and Code Explainer, supporting natural language to code generation across 10 programming languages and 5 database platforms including PostgreSQL, SQL Server, MySQL, Oracle, and SQLite. Built with a React frontend packaged inside an Electron desktop shell, NullCloud communicates directly with locally hosted large language models such as CodeLlama, Llama 3.2, and DeepSeek Coder through Ollama's REST API, enabling real-time code generation with zero API costs and no usage limits. The project includes two production database schemas embedded as structured context, a PostgreSQL recipe management system and a SQL Server zoo simulation, which are injected into model prompts to enable schema-aware query generation. Distributed as a cross-platform desktop application, the build pipeline leverages GitHub Actions to automatically compile and publish Mac and Linux installers on every release, while the Windows installer is packaged using electron-builder with NSIS targeting. The project demonstrates full-stack desktop application development with React and Electron, local LLM integration and prompt engineering, cross-platform CI/CD automation with GitHub Actions, and production-ready open source project structure including licensing, semantic versioning, and contributor documentation.

View Details Release

A Little About Me

I am currently a master's student at the University of South Florida actively looking for an internship. I enjoy this field of study a lot, I enjoy creating projects and my classes are really engaging to me. My coursework gives me valuable experience working with advanced database management, machine learning, and statistical analysis, and I am eager to build on my foundation. I am currently in two fellowship programs that are both used to train AI. These fellowships are basic in nature; however, they help me develop skills in data annotation and model training.

On a personal level, I love sports and video games. I spend my alone time watching sports and playing video games. I also love being with my friends and family, I often like to play cards with my friends. My favorite two sports are hockey and football, I admire the chaotic art that is hockey and the beutiful strategic warfare that is football. Some video games that I like are strategic games, where you have to make decisions to team or deck build.

Let's Connect!

Feel free to reach out!

LinkedIn GitHub Resume Email Me