Model and implement a real estate database: POC for a real estate agency to extract statistics by analyzing the real estate market
PostgreSQL
Docker
SQL
Excel (PowerQuery)
- Install and configure a database server
- Model a database
- Load data into a database
- Perform SQL queries to address business problems
Optimize data management for a shop: Sales analysis for a wine sales website
Python
Pandas
Plotly
- Search for outliers in data
- Multivariate analysis
- Generate graphs
Analyze sales for a bookstore: Activity report for an online sales company
R
tidyverse
kable
- Analysis of turnover (based on time)
- Sales analysis
- Correlation search (customer age / categories, etc.)
Conduct a study on drinking water: Creation of a dashboard for an NGO based on open data
Tableau
- Analysis of customer needs
- Implementation of a dashboard that meets expectations
- Generation of appropriate graphs
Market research production for a company wishing to export internationally
Plotly
ScikitLearn
Kmeans
PCA
- Open-data retrieval
- Data exploration
- Clustering
- Analysis report
Detection of counterfeit banknotes: Creation of a machine learning model to distinguish counterfeit banknotes
Numpy
ScikitLearn
Seaborn
- Modeling of numerical data (dimensions of banknotes)
- Training of a logistic regression model
Smart City competition: Analyze the trees in the city of Paris
Python
Seaborn
Plotly
- Data analysis (correlations, descriptive analysis)
- Processing of geographical data (plotting maps, heatmaps)
Prepare data for a public health organization: Cleaning a database and training a machine learning model
Python
Pandas
ScikitLearn
Flask
- Data cleaning
- Imputation of missing data
- Outlier analysis
- Analysis of variable relevance
- PCA / ANOVA
- Data visualization
- Creation of a prediction model for nutriscore based on nutritional data
- Creation of a deployed model presentation web app
Building a scoring model for a banking organization to determine if a client can be granted a credit
Python
ScikitLearn
- Descriptive analysis and cleaning of the dataset
- Feature engineering
- Sorting variables by feature importance
- Training multiple models with comparison
- GridSearch to optimize the hyperparameters of the best-performing model
- Cross-validation of models
Segmenting clients of a Brazilian e-commerce site and calculating the frequency of updating this segmentation
Python
ScikitLearn
- Descriptive analysis and data cleaning
- Feature engineering
- Utilization of clustering algorithms
- Simulating data aging for frequency calculation of model updates
Improving the AI product of a startup: Performing topic modeling on restaurant reviews and creating a photo classification model based on customer photos
Python
LDA
BERT
TensorFlow
Keras
- Data scraping
- Processing reviews using LDA
- Processing reviews using BERTopic
- Image classification using descriptor extraction with ORB
- Image classification using a CNN
Detecting Bad Buzz using Deep Learning in a set of tweets. Training a model for sentiment analysis on a tweet
Python
GenSim
Transformers
BERT
TensorFlow
Keras
Flask
- Text vectorization
- Lemmatization / Stemming
- Text embedding with Word2Vec, FastText, BERT
- Creation of an LSTM model with Keras
- Training the model and checking metrics
- Creation of a presentation web app
Contributing to the design of an autonomous car by modeling an image segmentation system
Python
Segmentation-models
TensorFlow
PyTorch
Flask
- Creation of TensorFlow Datasets
- Modeling of 3 models (Linknet, PSPNet, Unet)
- Training and comparison of models
- Creation of an API to call the model
- Creation of a demonstration web app
Developing a content recommendation application for a press website that aims to provide relevant recommendations to its users
Python
ScikitLearn
Surprise
Implicit
Flask
Azure Functions
- Modeling of two recommendation engines (Content-Based and Collaborative Filtering)
- Deployment of the model via Azure Functions
- Creation of a demonstration web app
Developing a chatbot for booking vacations with recognition of different variables and a monitoring strategy in production
Python
LUIS
Azure WebApp
Azure AppInsights
- Creation, Training, and Deployment of an Azure LUIS resource
- Development of the bot using the Microsoft framework
- Deployment of the bot on an Azure web app
- Production monitoring of the bot using Azure AppInsights
Preparing a presentation file for a Mobile application development project containing an AI functionality
Excel
Power Point
Azure sizing simulation
- Project presentation
- Financing, management of the budget forecast
- Agile/SCRUM methodology
- Project risk assessment
- AI ethics