Ophélie A.

| SALES ANALYSIS FOR AN E-COMMERCE SITE

Context: Production of the first monthly activity report after a change in company strategy (withdrawal of high-tech products from the catalog).

Tasks: Analysis and interpretation of key figures; Identification of the causes of the drop in turnover; Synthesis and formatting of data for reporting.

Technical Environment: Microsoft Excel, Power Query

| DESIGN AND USE OF A REAL ESTATE DATABASE

Context: A national network of real estate agencies wants to stand out from the competition by creating a database model to better predict property sale prices and better assist their clients.

Tasks: Preparation of the data dictionary, conceptual model, and normalized relational schema in 3NF; Import of the database into the database management system; SQL queries to analyze the market.

Technical Environment: SQL Power Architect, PostgreSQL

| STUDY OF MALNUTRITION IN THE WORLD

Context: The Food and Agriculture Organization of the United Nations (FAO), one of the UN bodies, aims to "help build a world free from hunger".

Tasks: Based on FAO data, a panorama of the state of malnutrition in the world for the year 2017 in particular.

Technical Environment: Jupyter Notebook, Python, Pandas

| OPTIMIZING THE DATA MANAGEMENT OF AN ONLINE WINE MERCHANT

Context: To manage resources, the company uses an ERP that is not connected to the online sales site, making inventory management complex. Pending a more centralized solution, manual reconciliation between the two databases is necessary.

Tasks: Reconciliation of the two exports; Dataset preparation; Analysis of turnover by product and overall; Search for outliers using several statistical tools.

Technical Environment: Jupyter Notebook, Python, Pandas, Numpy, Scipy, Seaborn

| ANALYSIS OF SALES FOR AN ONLINE BOOKSTORE

Context: Originally a physical bookstore, the company decided two years ago to launch an online sales site. You are brought in because the company wants to take stock after two years of operation, and be able to analyze its strengths, weaknesses, customer behavior, etc.

Tasks: Analysis of references sold, top and bottom performers by product and category; Analysis of turnover, its evolution, and its distribution according to customer base; Analysis of customer profiles and purchasing behavior (variable dependency tests).

Technical Environment: Jupyter Notebook, Python, Pandas, Numpy, Scipy, Statsmodels, Datetime, Seaborn, Matplotlib

| ANALYSIS OF GENDER EQUALITY INDICATORS

Context: A consulting firm already has over 150 employees and is in full development. As every year before March 1st, companies with at least 50 employees must calculate and publish their gender equality index on their website.

Tasks: Study of directives and selection of indicators to monitor; Preparation of data from the HRIS in compliance with GDPR; Automation of the diagnostic report creation with the KNIME workflow.

Technical Environment: KNIME

| STUDY OF ACCESS TO DRINKING WATER IN THE WORLD

Context: The NGO DWFA (Drinking Water For All) aims to provide access to drinking water for everyone. For a funding request, a dashboard needs to be created presenting a global view of access to drinking water worldwide to help choose a target country.

Tasks: Data cleaning via Tableau Prep Builder; Needs analysis to formulate analytical questions; Data formatting; Creation of a dashboard presenting a global view of access to drinking water worldwide.

Technical Environment: Tableau Prep Builder, Tableau Desktop, Tableau Public

| STUDY OF THE GLOBAL MARKET TO DEFINE AN EXPORT ZONE

Context: A poultry producer wants to expand internationally. Currently, all continents and countries are possible. Identify groups of countries where export is possible, determine the commonalities of these country groups, then make a recommendation.

Tasks: Choice of variables to keep; Variable reduction using Principal Component Analysis; Individual reduction using Hierarchical Ascending Classification; Simple clustering using the k-means method.

Technical Environment: Jupyter Notebook, Python, Pandas, Numpy, Scipy, Scikit-learn, Matplotlib, Seaborn

| CREATION OF A COUNTERFEIT BANKNOTE DETECTION ALGORITHM

Context: Using a sample file containing 1,500 banknotes and 6 of their geometric characteristics, implement an algorithm capable of automatically identifying counterfeit banknotes.

Tasks: Dataset cleaning and imputation of missing values by multiple linear regression; Comparison of the performance metrics of two prediction methods (logistic regression and k-means); Algorithm implementation.

Technical Environment: Jupyter Notebook, Python, Pandas, Numpy, Scipy, Statsmodel, Scikit-learn, Matplotlib, Seaborn

Reading, analysis, and simplification of the latest research work published on early childhood development; writing reports.

- Internal Communication: Implementation of internal communication processes.

- Community Management: Management of the startup's social networks.

- B2B and B2C Communication: Writing, design, and production monitoring of communication materials.

- Financial Communication: Business Plan writing, elevator pitch...