Ralph Touma's playable résumé
- ←→ Move
- Shift Run
- Space Jump
- ↑↓ Climb · Pipes · Crouch
- P Pause
- T Read as text
- M Mute
- F Full screen
Ralph Touma, machine learning engineer and data scientist in Paris
SUPER FOLIO · text version
Ralph Touma is a machine learning engineer and data scientist based in Paris, France, working on federated learning, forecasting models and retrieval-augmented generation (RAG). Ralph turns ideas into working systems: clean data pipelines, models evaluated honestly, and products that keep performing once they meet real users.
SUPER FOLIO is Ralph's résumé as an original 8-bit platformer: each of its six checkpoints opens a chapter, and the castle at the end holds the contact card. Here is every chapter as text, for anyone who would rather read than play.
- Based in
- Paris, France
- Focus
- Federated learning, forecasting, RAG
- Certified
- 3× AWS, including Generative AI Developer (Professional)
- Languages
- French, English, Arabic
Checkpoint 1/6 · Player select
Profile
Machine learning engineer and data scientist in Paris, France. Ralph's work spans machine learning research, forecasting, AI assistants and data products for real clients. Whatever the scale, Ralph cares about the same things: understanding the data, measuring what matters, and shipping something that holds up.
Special moves
- End-to-end machine learning pipelines
- Forecasting models
- Data preprocessing
- Experiments and validation
- Analytics dashboards
Languages
Speaks French, English and Arabic.
Checkpoint 2/6 · Training arc
Education and certifications
Master's in Data Science, Saint Joseph University of Beirut
Bachelor's in Computer Science, Notre Dame University
Trophies unlocked!
Certifications
-
April 2026
AWS Certified Generative AI Developer – Professional
Early adopter
Verify on Credly -
March 2025
AWS Certified Machine Learning Engineer – Associate
Verify on Credly -
January 2025
AWS Certified AI Practitioner
Verify on Credly
Checkpoints 3–4/6 · Main quests
Experience
Web Developer, BYND Network
Quest in progress
- Shipped full-stack websites for 15+ agency clients, from front end and back end to database and deployment.
- Built and customised Shopify storefronts with Liquid, JavaScript and the Shopify API.
- Developed custom Shopify apps with Node.js to extend store functionality.
- Built custom web apps from scratch with React and Vue for clients beyond Shopify.
- Improved a client's conversion by 30% through visitor behaviour analysis.
Case study: commerce engineering and analytics at BYND Network
Machine Learning Research Intern, Green-Communications
Quest complete
- Built ML policies and a forward model (R² ≈ 0.97) to configure wireless mesh network routing for Green Communications, replacing an approach that could not work.
- Tested configurations on 10, 28 and 104-node networks and fixed the 104-node shortfall with domain-informed constraints.
- Implemented a three-layer clustered federated learning framework for soil-moisture sensors, simulated on 80 sensors.
- Reached RMSE 2.43 after 12 readings in cold-start tests, against 3.18 for local models and about 3.2 for a single global model.
- Co-authored a paper submitted in 2026: “A Three-Layer Game-Theoretic Framework for Soil-Moisture Clustered Federated Learning”.
Case study: machine learning research at Green-Communications
Freelance Full-Stack & AI Engineer, Independent
Quest in progress
- Designs, builds and ships full-stack web applications and websites end to end, from scoping and architecture to front end, back end, database, deployment and handover.
- Builds AI systems and autonomous agents for clients: LLM-powered assistants, retrieval-augmented generation over private documents, tool-using agents and workflow automation wired into existing products.
- Owns every client relationship directly, turning loose business goals into scoped technical plans, production code and measurable results.
- Works across React.js, Vue.js, Node.js, Express, Python, SQL and MongoDB, with models and agents deployed on AWS and Azure.
Backend Developer Intern, Eurisko
Quest complete
- Engineered back-end services in Node.js and Express.js, designing the server-side logic and endpoints that powered the front end.
- Optimised application performance and responsiveness across the request cycle.
- Modelled and managed MongoDB (NoSQL) data storage for application data.
- Collaborated with front-end developers in a Git-based team workflow.
Search Engine Evaluator & Data Collector, Appen
Quest complete
- Evaluated the accuracy and relevance of search results returned by virtual assistants including Siri, Alexa and Google Assistant.
- Extracted and classified business information from media sources, supporting structured data optimisation for AI training datasets.
Checkpoint 5/6 · Boss battles
Projects
Enterprise RAG assistant for technical documents
Boss: hallucinations
- Ingests PDFs and internal documents, then chunks, embeds and stores them as vectors for semantic retrieval.
- Hybrid retrieval, semantic search plus lexical matching, then relevance ranking, so exact terms like part numbers and function names still find their passage.
- Answers grounded in citations to the source documents, so a reader can check an answer rather than trust it.
Privacy-aware federated weather forecasting
Boss: non-IID station data
- Next-day forecasts across city-level clients, built on Flower's simulation framework: each city trains on its own station records and shares model updates rather than raw observations.
- Personalized for each city, so every client keeps a model adapted to its own conditions.
- Benchmarked centralized, local-only, FedAvg and personalized federated models on the same data.
- Per-client and global error across rounds, with custom metric aggregation.
Checkpoint 6/6 · Inventory
Skills
Programming
- Python
- R
- SQL
- JavaScript
- TypeScript
- Java
- C++
- C#
Data & ML
- pandas
- NumPy
- scikit-learn
- TensorFlow
- PyTorch
- Hugging Face
- Amazon SageMaker
- Jupyter
Cloud
- AWS
- Microsoft Azure
Visualization & BI
- Power BI
- Looker
- Excel
- Matplotlib
- Seaborn
Databases
- MySQL
- MongoDB
Web & APIs
- React
- Vue.js
- Node.js
- Express.js
- REST APIs
- HTML
- CSS
In the game, each coin a block pays out floats up as the next of these skills instead of its score.
The castle · Player 2 wanted
Contact
Research collaborations, machine learning roles in Paris, or a product that needs building: email is the fastest way to reach Ralph.
About SUPER FOLIO
SUPER FOLIO is an original 8-bit side-scrolling platformer that Ralph Touma built from scratch in vanilla JavaScript and HTML5 canvas, with no game engine, libraries or build step. Its hero, Rivet, is a mechanic in a yellow hard hat. The pixel art is painted from character grids when the page loads, and the chiptune music and sound effects are synthesized live in the browser from pulse, triangle and noise channels.
It plays with classic momentum: acceleration, skidding, jumps that go higher at speed and the longer the button is held, and a camera that only scrolls forward, all on a fixed 60 Hz simulation. Each checkpoint freezes the level and opens a chapter of the résumé. Along the way there are power-up gears, kickable shells, a hidden 1-UP, a pipe down to a coin room, a flagpole and a castle.
Questions about SUPER FOLIO
What is SUPER FOLIO?
SUPER FOLIO is the playable résumé of Ralph Touma, a machine learning engineer and data scientist in Paris. It is an original 8-bit side-scrolling platformer: each of its six checkpoints opens a chapter of the résumé, coin blocks pay out skills instead of points, and the castle at the end holds the contact card.
Who made SUPER FOLIO?
Ralph Touma built it from scratch in vanilla JavaScript and HTML5 canvas, with no game engine, libraries or build step. The hero, Rivet, is a mechanic in a yellow hard hat. The pixel art, chiptune music and sound effects are all original, and the music is synthesized live in the browser.
Can I read the résumé without playing?
Yes. Pick READ AS TEXT on the title screen: everything the game shows is on this page as text, chapter by chapter, and the full portfolio with case studies is at ralphtouma.com.
How do you play SUPER FOLIO?
On a computer, move with the arrow keys or A and D, run with Shift, and jump with Space (hold it to jump higher). The up and down arrows climb vines and enter pipes, P pauses, M mutes and F plays full screen. On a phone or tablet, hold it sideways: the game fills the screen, with an on-screen D-pad and an A button.