Professional experiences

■ 01.2024 - now ■ Data Consultant at Capgemini Invent

○ Generative AI Projects:

• Focused on exploring and implementing cutting-edge, State-of-the-Art methodologies.

• Developed virtual assistant software for Question-Answering systems, specializing in medical queries.

• Built virtual assistant software to process complex audit documents and reports, incorporating features such as question-answering, summarization, template-based proofreading, and document comparison.

○ Technical Keywords: Chatbot, Retrieval-Augmented Generation, Large Language Models, Generative AI, Unstructured, PyMuPDF, Streamlit, GCP (Document AI), Azure (AI Search, OpenAI, AI Services etc)

○ Fields: Airline Industry, Medical Industry, R&D

■ 04.2021 - 12.2023 ■ Data Scientist/Researcher at Quantmetry

○ R&D in Generative AI:

• Developed Qolmat, an open-source library for imputing tabular and time-series data using generative models.

• Implemented data augmentation techniques to improve service evaluation in the MAPIE project and contributed to quality evaluation of generated data.

• Engineered generative models for synthetic designs with constraints and performed quality evaluations of generated data.

○ Technical Keywords: Data Augmentation, Generative Models, Generative Adversarial Network, Variational Auto-Encoder, Diffusion Models, Transformers, Trusted AI, Generative AI, NLP, Tabular Data, Time Series

○ Fields: Automotive Industry, R&D

■ 10.2017 - 02.2021 ■ Generative Probabilistic Alignment Models for Words and Subwords ○ NLP (Utility)

○ Researched alignment models for low-resource languages, focusing on subword tokenization and variational inference techniques.

○ Doctoral Project: 3 years in Laboratoire d'Informatique pour la Mécanique et les Sciences de l'Ingénieur (LIMSI)

○ Technical Keywords: Low-resource Languages, Variational Auto-Encoder, Subword Tokenization, Expectation-Maximization, Hidden Markov Models, Convolutional Neural Networks, Long Short-Term Memory Networks, SentencePiece, Giza++, Fastalign, Tensorflow, Pytorch, Python, Slurm

○ Thesis.fr

■ 03.2017 - 09.2017 ■ Neural Machine Translation ○ NLP (Utility)

○ Developed GAN-based Neural Machine Translation systems and explored reinforcement learning applications for improved translation quality.

○ Internship: 6 months in Laboratoire d'Informatique pour la Mécanique et les Sciences de l'Ingénieur (LIMSI)

○ Technical Keywords: Generative Adversarial Network, Reinforcement Learning, Convolutional Neural Network, Recurrent Neural Network, Python

○ Github: GAN-NMT 

■ 09.2016 - 11.2016 ■ Axa Data Challenge ○ Data Challenge of DaSciM

○ Developed predictive models to address concept drift in time-series data.

○ Scholarly Project: 3 months in École Polytechnique (X)

○ Technical Keywords: Concept Drift, Tree Regressors, Random Forest Regressors, Gradient boosting regressors, Autoregressive models, Python, Scikit-Learn, Anaconda

■ 01.2016 - 03.2016 ■ Brain MRI Segmentation ○ Software Development (Medical)

○ Developed software for segmenting brain MRIs using 3D slicer and ITK.

○ Internship: 3 months in Kyoto Institute of Technology (KIT)

○ Technical Keywords: C++, 3D Slicer, ITK

○ Youtube: Demo

■ 06.2015 - 08.2015 ■ Project KList ○ iOS Development (Entertainment)

○ Internship: 3 months in NinePoints Co. Ltd 

○ Technical Keywords: Swift, ObjectiveC, Xcode, Realm, Alamofire 

○ Appstore: Klist

■ 02.2015 – 06.2015 ■ Project 3D Movie ○ 3D Development (Entertainment) 

○ Graduation Project: 4 months in Arena Multimedia Institute (AMI)

○ Technical Keywords: Maya 3D 

○ Youtube: The Ant and The Grasshopper

■ Machine Learning

○ Large Language Models, Retrieval-Augmented Generation

○ Variational Auto-Encoder, Generative Adversarial Networks, Diffusion Models

○ Convolutional Neural Networks, Recurrent Neural Networks, Transformers

○Support Vector Machines, Back-propagation, Naive Bayes Classifier, Linear Regression, K-Nearest Neighbor, Decision Trees 

○ Expectation-Maximization, DBSCAN, K-Means, Hierarchical Clustering 

○ Boosting Algorithms, XGBoost, LightGBM

○Pytorch, Theano, TensorFlow, Scikit-Learn, LibSVM, Weka, MLFlow, FastAPI, Slurm, AWS, Google Cloud, Azure, Jupyter Lab, DeepNote

○ Transformers, Spacy, SentencePiece, Spark, Giza++, Fastalign, Simalign, Eflomal

■ Software development ○ Python, Scala, R, Matlab ○ Java (J2EE), C/C++, C# ○ OpenCV, Qt ○ Lisp, Prolog, Jess ○ Netbean, Eclipse, Visual Studio, Anaconda, Spyder

■ Mobile development ○ Objective C, Swift, React ○ iOS, Android ○ XCode, Android Studio

■ Web development ○ HTML5, CSS3, JavaScript ○ PHP ○ jQuery ○ Dreamweaver ○ Adobe Flash

■ Multimedia development ○ Illustrator, CorelDraw ○ InDesign ○ Photoshop, Lightroom ○ Premiere Pro, After Effects ○ 3ds Max, Maya 3D

■ DB Management ○ SQL ○ MySQL, Microsoft SQL Server ○ Microsoft Access

■ Architectural Design ○ UML

■ Project management ○ Project planning and roadmapping, Agile methodologies, Documentation development ○ Trello, Microsoft Office Suite