Abdessalam DAI

Skilled data scientist with 2 years of experience at Amadeus (leading travel-tech provider), building large-scale data pipelines and machine learning models for high volume flight data. Passionate about transforming data into actionable insights that enhance user experience, optimize operational efficiency, and boost revenue.

Predictive Modeling of Apartment Rent Prices in the US

• Engineered new features from external sources (e.g., population, average salary) to enhance predictive performance.
• Trained various models including XGBoost, CatBoost, LGBM, and stacking ensembles.
• Used MLflow for experiment tracking and YData Profiling for data quality assessment and model diagnostics.
• Built a Streamlit dashboard with interactive pricing maps and SHAP-based feature explanations.

Data-Driven Analysis of Collaborative Workflows Using Tactile Interaction

• Analyzed user interaction logs from interactive touch tables to identify collaborative behavior patterns.
• Applied data science methods to extract, clean, and structure behavioral data from raw logs.
• Developed a Python simulation tool to replay interaction sessions and visualize user behavior.
• Enabled research teams to interpret patterns through data-driven visual outputs.

Visual SFSD – web app for visualising fundamental file & data structures (link: vsfsd.estin.dz)

• Full-stack build (Python / Django, D3.js, Tailwind CSS) in a 5 student team over 3 months.
• Developed an interactive interface that demonstrates core file-system structures and algorithms for computer-science students.

Recruitment Platform – job-matching website built with React (front-end) and Express.js (back-end)

• Implemented role-based workflows for admins, recruiters, and candidates: posting offers, searching listings, and submitting applications.

Data Analyst Nanodegree - Udacity

Machine Learning - Coursera, Andrew Ng