Portfolio

IA Projects

Prediction of Wine Quality
I conducted a Machine Learning project focused on predicting wine quality based on composition analysis. The analysis encompassed statistical exploration, including data types, missing values, and correlations. For prediction, I applied both regression, using Linear Regression and model comparisons, and classification, employing methods like K-NN, SVM, Decision Trees, and Bayes Classifier. The results underwent visual representation, emphasizing scatter plots and decision trees. The theoretical aspect involved providing algorithmic formalism, detailing parameters, and concluding with method comparisons. The report not only discussed the motivations behind the chosen methods but also analyzed results and proposed potential enhancements, such as combining methods for improved wine quality prediction.

Clustering Of Mobile Price on Today’s Market
I led a clustering project using market mobile prices to estimate our (fake) company’s mobile prices. Employing techniques like PCA, KMeans, t-SNE, and k-Medoids, I aimed to create new cluster labels aligning with the ‘price_range’ target variable. The analysis involved dimensionality reduction with PCA, clustering with KMeans and k-Medoids, and visualization using t-SNE. The report highlights the efficacy of these methods in creating meaningful price clusters, providing insights for our pricing strategy.

Estimacloud with OnePoint
Role : Product Owner and Python Developper
I spearheaded a pricing tool project estimating cloud infrastructure costs using Flutter for the web interface. Elasticsearch and Logstash were employed to retrieve real-time cloud service prices, while Flask and the Random Forest machine learning algorithm were utilized for invoice preparation. This comprehensive approach integrates user-friendly design, dynamic pricing data extraction, and machine learning for accurate billing, providing an efficient solution for estimating cloud infrastructure expenses. ( No code because this tool is now used by OnePoint)

Visualization Projects

Japan Netflix Preferences
In my Tableau-based visualization project on a Netflix database, I addressed four pivotal questions. I utilized visual insights to recommend the optimal streaming platform by analyzing genre popularity, user ratings, and viewer demographics. The decision between a movie or TV show adaptation was informed by visualizing data on runtime, genre trends, and audience engagement. To identify the best TV show or movie type for adaptation, I employed data visualizations highlighting genre preferences, viewer ratings, and popular themes. Lastly, I recommended an appropriate director based on visual analytics, considering factors such as viewer ratings, awards, and genre expertise. This comprehensive approach provides actionable insights for strategic decision-making in content adaptation.

Accenture’s Project
In a group project simulating the role of a data consultant at Accenture, our focus was on addressing key challenges for Terrace Solution, a leading online sales platform expanding into the Brazilian market. With an annual growth rate exceeding 11%, Terrace sought insights into this new market’s dynamics. Our project team was tasked with providing valuable insights tailored to specific stakeholders. The Brazilian General Manager desired a macroscopic view of sales and conversion rates from email campaigns. Simultaneously, the Business Unit Manager sought detailed information on sales (outstanding/paid) and the status of order deliveries. Through data analytics and visualization, we aimed to deliver actionable insights to support Terrace’s strategic decisions and operational efficiency in the Brazilian market.

Database Management Projects

Avanade’s Project
This Avanade’s Project has for objective to enhance the international positioning of EFREI PARIS. For this, the school signed partnerships with several major schools and universities around the globe. it is about having a reporting tool that will allow him to follow the students during their schooling but also in their professional life through the alumni association of which he remains the sponsor. Using a robust data model and management KPIs, the tool should also allow to follow various things like course programming, monitoring the availability of teachers, end of year assessments and exams, payment of tuition fees or salaries of school staffs.

Theatre Organization
Theater is the art of writing and performing plays but it’s also the structure which have rows seats, from where people can watch a performance. In fact, theaters are less and less watched by people due to all other activities deemed more interesting.So how can all theatrical companies still survive in these years of digital dominance? To face this problem some theatrical companies had the idea of making a common management of all their shows, so their theaters can welcome many other theatrical companies and their staff can play in other places and not only on their theater respectively.