Data Scientist / AI Engineer

Technical Skills:

Programming Languages: Python, SQL, NoSQL, Bash, Java, C++, Matlab

Libraries and Frameworks: Matplotlib, Seaborn, Plotly, Pandas, Numpy, SciPy, Scikit-Learn, Prophet, Darts, Pytorch, Tensorflow, Keras, Spacy, nltk, Hugging Face, FastAPI, Streamlit

Cloud Computing, Containerization and MLOps: AWS, OVH, Docker, CI/CD pipelines, unit testing

Data Viz tools: Excel, Tableau, Google Analytics, Dataiku, Dash

Education

M.S. (bac +5), Computer Science (Data Science and Analytics track) EPITA (April 2023)
B.S. (bac +3), Computer and Communication Engineering LIU (August 2021)

Work Experience

AI and Data Consultant @ Quant AI Lab (January 2024 - Present)

Development of a data pipeline to collect ESG (Environmental, Social, and Governance) information from data on 30,000 companies in Spain.
Utilization of the database using GDELT data and GCP BigQuery to query and process news data at scale.
Application of HuggingFace’s ESG classification model and BERT models for scoring on Vertex AI Workbench and GCE, followed by the development of a results visualization on a Streamlit dashboard to generate insights.
Orchestration and industrialization of the pipeline on GCP via Cloud Composer to schedule/monitor workflows, and Dataflow to execute distributed processing (data ingestion, transformation) at scale.
Development of a chatbot capable of answering questions based on PDF documents uploaded by the client.
Implementation of the RAG system to improve the chatbot’s accuracy in understanding and answering questions.
Integration of LangChain and the OpenAI API with the GPT model to facilitate conversational responses and knowledge extraction from PDFs.
Implementation of an agent-based workflow with LangGraph to orchestrate the chatbot’s reasoning (query routing, retrieval steps, context verification, rephrasing if necessary) and ensure reliable responses via state graph execution.
Developed a warehouse box monitoring system that uses computer vision techniques to estimate the volume of each box, thereby supporting inventory management.
Implemented VILD for object detection and InstantMesh for 3D reconstruction to analyze and map the warehouse environment using a collaborative approach.
Collaborated with the team to integrate MediaPipe, enabling tracking and localization of the employee scanning the boxes, ensuring accurate data capture and smooth workflow management.

AI Engineer @ Helean (Octobre 2022 - April 2023)

Conducted data analysis on huge Time Series client financial datasets to gain insights and inform decision-making processes.
Extracted relevant information from large datasets to identify patterns and trends; it includes data mining.
Contributed to the development of machine and deep learning models, as well as the creation of a basic recommender system model for customers (event description analysis using NLP techniques and word embeddings using Spacy, nltk, Camembert, flair).
Worked with PyTorch, Scikit-learn, and TensorBoard for model development and optimization.
Contributed to data processing pipelines for data ingestion, transformation and feature engineering.
Used data visualization tools (Jupyter, Pycharm, VScode, Matplotlib, Seaborn, Pandas and Excel) to present findings and communicate results to the founders