About Julien
Ingénieur en Informatique spécialité Big Data, je suis passionné de Data Science, d’intelligence artificielle et de nouvelles technologies. Ces domaines offrent la possibilité de toucher à de nombreux secteur d’activités et problématiques métiers, c’est pour cela que je me suis inscrit sur malt afin de pouvoir réaliser des missions captivantes et mettre mes compétences à votre service pour vous accompagner dans vos nouveaux défis !
Linkedin : Julien Roumagnac
French
Native or bilingual
English
Fluent
Can work on-site
Lyon (up to 50km)
Experience
- UbisoftData ScientistVIDEO GAMES AND ANIMATIONSeptember 2022 - Today (3 years and 9 months)Lyon, FranceWorking on In Game player exeprience personalization & recommandation
- AVISIAConsultant Data scientistDIGITAL AND ITMarch 2020 - September 2022 (2 years and 6 months)Lyon, FranceAvisia internal roles :○ Lead of a computer vision R&D project : developement of a web application to detect & localize specifics objects in a picture using Deep Learning○ GCP Machine Learning Engineer instructor : co-creation of a study path for the certification (main ressources, milestones, trainings necessary to be ready for), co-animation of training sessions, mentoring participants during the process○ Contributor : Particpation to internal streams to present technical subject, feedbacks etc and Internal articles writingWithin the DATA & AI Lab of a major player in fashion and luxuryMissions :Scoring Categories of Interest (Machine Learning):○ Development of Machine Learning algorithms for scoring customer appetence to different defined product categories○ Analysis of the model with XAI methods (AI Interpretation) to allow business teams to understand the algorithm's decisions○ Development of a model training/scoring/monitoring pipeline○ Analysis and monitoring of prediction results, comparing it to traditional methods.○ Automation of this pipelineEmailing optimization (Deep Learning):○ Implemented a product recommendation algorithm for the customer's next purchase for an email prioritization context to select the most relevant emails for each customer○ Development of the algorithm based on an LSTM architecture○ Development of Dataiku training & scoring flows○ Development of a dash web app to allow business teams to easily interact with the algorithm○ Development of the first Dataiku use case of the company => participation in various tests and reflection on the implementation, use and governance of the tool for future projectsElaboration of ML Engineering / ML Ops best practices:○ In collaboration with the Lead ML Engineer :○ Realization of workshops to define the ML Ops needs of the team.○ Definition of the best practices to be implemented○ Design of the generic ML ops architecture for the Lab's ML projects○ Application of this architecture to my project Categories of interest○ Coaching other Data Scientists to implement this architecture on their projectsAnalysis and specific developments :○ Analysis and understanding of the customer base○ Study of available data sources (customer data, web browsing and social networks)○ Statistical analysis and cross-referencing of data sources○ Data processing and cleaning, creation of DataMarts dedicated to analysis○ Definition of KPIs in collaboration with the Business Analyst○ Development of customer segmentation and analysis methods based on KPIs○ Development of a Machine Learning algorithm for scoring the appetence of customers for the purchase of a 1st "ready to wear" purchase during the next month.
- VOLPI IMMOBILIERData ScientistREAL ESTATEAugust 2020 - September 2020 (1 month)Lyon, France#Python #MachineLearning #Scikit-learn #Xgboost #pandas #API #Flask #Pickle-> Création d’un estimateur de valeur de biens immobiliers :• Etude de la faisabilité et cadrage du projet• Analyse des sources de données• Analyse statistique et traitement des données• Construction de la base d’études• Développement et optimisation d´un algorithme de régression• Compression de l’algorithme pour son utilisation en production• Mise en production de l’algorithme d’estimation via une API Flask• Création d’une routine permettant de mettre à jour l’algorithme régulièrement pour prendre en compte les nouvelles données disponibles
Reviews
Recommendations
Be the first to recommend Julien
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
- Ingénieur en informatique, spécialité Big DataPolytech Montpellier2020