Freelancer profile translated to English.
Back to original languageAbout Erwan
Curious and passionate Data Scientist with over 6 years of experience, I have acquired solid expertise in data analysis, the design and deployment of Machine Learning models, and the creation of solutions for their exploitation by business teams.
My professional background has led me to work in numerous sectors (marketing, communication, e-commerce, etc.) and to diversify my skills in Python development and Machine/Deep Learning technologies, with a keen eye on new innovations in the world of artificial intelligence (GenAI, LLM, RAG, etc.) and their application in my latest projects.
These experiences have also allowed me to be involved in all phases of the development of these solutions, from data extraction and manipulation to model deployment and development cycle management.
I am aware of the challenges involved in applying these solutions in a business project and am committed to working at all levels of their development to ensure an optimal final result and maximize ROI for decision-making teams.
French
Native or bilingual
English
Fluent
Spanish
Conversational
Can work on-site
Bordeaux (up to 50km), Paris (up to 10km)
Experience
- Erwan VautierData ScientistCONSULTING AND AUDITSDecember 2021 - Today (4 years and 6 months)Bordeaux, FranceE-commerce platform specializing in online tool sales (1 year):
- Automatic classification of tooling products into about fifty categories based on product descriptions and images (NLP + Computer Vision). Testing of different methodologies and prediction models.
- Creation of a product attribute detection model from descriptions using LLMs with various open-source or proprietary models (Mistral, LLama, OpenAI).
- Creation of a Flask API to call different models from requests on the PrestaShop module (automatic detection of a category and attributes during product import on the e-commerce site).
- Management of the development cycle for various projects (Git, Docker, Kubernetes, CI/CD).
Global communication and advertising group (6 months):- Encoding of products based on their descriptions and images (NLP + image feature extraction) and development of Machine Learning models based on the obtained vectors (multimodal): segmentation, clustering, scoring. Use of the CLIP model to extract semantic features from images and build multimodal vectors.
- Integration of various Google Cloud Vision (GCV) APIs during the processing phase to enrich textual descriptors of images (object detection, text detection, etc.).
- Management of the development cycle (Git, Docker) and deployment of models in an AWS cloud environment (EC2, S3, SageMaker, Lambda).
Various projects:- Creation of web data retrieval tools in Python (requests, BeautifulSoup, Selenium, Scrapy, etc.).
- Development of Windows or macOS applications (Kivy, Tkinter) to utilize various tools or models (e.g., LinkedIn scraping and lead qualification).
- Creation of an embedded system on Raspberry Pi for user listening, communication with a fine-tuned model hosted by OpenAI, and voice output.
- INEOXData ScientistCONSULTING AND AUDITSOctober 2019 - November 2021 (2 years and 1 month)Paris, FranceLeading online real estate platform (1 year):
- Processing and analysis of real estate data for media purposes. Creation, management, and deployment of statistical barometers and dataviz exploited by business teams (Tableau).
- Development of Machine Learning models for various tasks (detection of duplicate listings from images and descriptions, prediction model for housing prices based on characteristics, etc.)
- Utilization of time series models (ARIMA, SARIMA) to predict the evolution of rent and seasonal rental prices.
National network of optical stores (6 months):- Audit of data environment quality, analysis of transactional and customer data, descriptive report of the current state (missing values, distribution, statistical links), and development of solutions to implement.
- Creation of Machine Learning models (customer segmentation, sales and churn/anti-churn prediction model, item-item and user-user recommendation system), automation, and marketing activation (online platform, personalized email campaigns, etc.)
Various projects:- Management and manipulation of databases, extraction and processing, creation and deployment of automated data flows.
- Processing and analysis of data for marketing needs, descriptive analysis of customer data, sales, and traffic, and reporting of analyses to business teams.
- Development of Machine/Deep Learning models for various companies: Customer segmentation, sales/churn prediction models, etc.
- TheContilleryData ScientistCONSULTING AND AUDITSApril 2018 - October 2019 (1 year and 6 months)Paris, France
- Design, development, and deployment of several image processing models (CV) to enhance visual communication for brands on social media. Clustering based on features extracted from pre-trained models like ResNet (deterministic KMeans, HAC), automatic image classification, object detection and extraction (Mask R-CNN), and visual performance prediction model.
- Exploration and research of new methods to overcome challenges inherent to certain categories (object extraction, distance research, new models, etc.)
- Development of Python tools for retrieving engagement data related to images on social media (WebScraping/Crawling).
- Design of web applications (Flask) for visualizing, exploiting, and validating models (clustering, engagement score, image similarity, etc.)
- Deployment of data flow extraction and control tools. Management of the development cycle for various projects (Git, Docker, CI/CD)
Recommendations
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- Engineering degree, Mathematics and StatisticsEcole nationale de la Statistique et de l'Analyse de l'Information2018Statistiques et Analyse de l'information, Data Science, Machine/Deep Learning, Modélisation. Python, R, SAS.
- Bachelor's degree in Applied EconomicsUniversité Paris Dauphine2017Licence en économie appliquée en double diplôme avec l’ ENSAI