Freelancer profile translated to English.
Back to original languageAbout Alexandre
đšâđ»Freelance Data Scientist since 2021. Currently holding a Data Scientist position within the Datalab of SociĂ©tĂ© GĂ©nĂ©rale CIB.
đ Double degree from ENSTA Paris & Institut Polytechnique de Paris. Preparatory classes PCSI / PSI* at LycĂ©e Pierre de Fermat.
đ§ Specialties: NLP, Computer Vision, Time Series, Deep Learning, OCR, Machine Learning, LLM.
đŻ Skills:
- Information extraction from documents
- Document OCR
- Image or text classification
- Image segmentation
- Sentiment Analysis
- Q&A
- Text data mining (topic identification / parsing / document classification)
- Anomaly detection
- Clustering
- Predictive analysis
- Data Visualization
- Dashboards
- Interfaces
- Decision support
đ Methodology:
- Scoping: needs analysis, identification of challenges, constraints, and added value through AI. Data audit.
- Establishment of a milestone-based schedule.
- Modeling using classic Machine Learning or Deep Learning depending on the case. Proof of Concept (PoC) production.
- Iterative model optimization.
- Deployment.
- Maintenance and support.
đ ïž Stack:
- Python
- Scikit-learn / Numpy / Scipy
- Pytorch / Tensorflow / Keras
- NLTK / YOLO / PaddleOCR
- HuggingFace / LangChain
- Git
- Docker / Kubernetes
- Flask / FastAPI
- Azure Cognitive Services / Google Cloud
đ Recent achievements: Information extraction from invoices, identity document classification, finetuning of an OCR neural network for low-resolution documents, NER classification.
đ Thank you for visiting my profile, and I look forward to discussing your projects!
French
Native or bilingual
English
Fluent
Spanish
Conversational
Remote only
Primarily works remotely
Experience
- Groupe HLDDevelopment of an information extraction model from client documents for a KYC project.BANKING AND INSURANCEJanuary 2024 - February 2024 (1 month)Paris, FranceExtraction of specific information obtained by mining document bundles.Technically, this project required:
- Imageclassificationto filter relevant documents.
*OCRfinetuned on the specific corpus.- Client **matching**.
- Extraction of key information using a **multimodal text + image model**.
- Validation rules.
- Deployment viaDocker
- Société Générale - CIBData ScientistBANKING AND INSURANCEOctober 2022 - Today (3 years and 8 months)Paris, France- Anomaly detection based on Graph Neural Networks- Text classification
- LotomateData ScientistSOFTWARE PUBLISHINGOctober 2023 - January 2024 (3 months)Paris, France
- Information extraction from various documents (invoices, CVs, CAF certificates, tax notices, ...) using a multimodal model (text & image) and leveraging document structure analysis (Layout Analysis).
- Automation of training by integrating the model into Wirk's microservices (connecting to the database to retrieve annotated training documents).
- Inference module to obtain predictions directly in the annotation interface.
Reviews
Recommendations
Be the first to recommend Alexandre
Help this freelancer shine by sharing your experience working together.
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 from ENSTA Paris - Specialization AIENSTA Paris2022Double diplĂŽme d'ingĂ©nieur ENSTA Paris & Institut Polytechnique de Paris en spĂ©cialitĂ© IA. Formation sur l'IA mĂȘlant thĂ©orie (MathĂ©matiques AppliquĂ©es, Optimisation, ProbabilitĂ©s et Statistiques) et la pratique (nombreux projets en Machine Learning et Deep Learning). CPGE en 2 ans : PCSI / PSI* au lycĂ©e Pierre-de-Fermat. BaccalaurĂ©at spĂ©cialitĂ© Maths avec mention TrĂšs Bien (18,2 de moyenne)
- Master 2 Data & AIInstitut Polytechnique de Paris2022Master 2 Data & IA avec une spécialisation en Machine Learning et Deep Learning. Cette formation est ponctuée de nombreux projets dont voici quelques exemples : - Classification de texte et Named Entity Recognition (NER) [NLP] - Classification et Génération d'images (GANs) [Computer Vision] - Séparation de pistes d'un morceau de musique [MIR] - Recommandation d'articles pour site de e-commerce [SystÚme de recommandation] - Apprentissage par Renforcement
Certifications
- Natural Language Processing with Attention ModelsCoursera2023
- Natural Language Processing with Classification and Vector SpacesCoursera2023