Julien -heiduk

- Cold Start problem - Multi-modal similarity between items - Recommendations personalized - A/B tests - LLM for compatibility between items - GNN apply to recommendation system

I help aspiring Data Scientist & ML Engineer to achieve their goals. The satisfaction rate of the 32 learners who have rated me and whom I have already coached is 4.94/5.

Time Series Classification & API

- End-to-end ML-project with Dataiku for Predicitve SEO - Text generation for products description, FAQ, article generation - Integration of Data Sources (From API & scraping) and creation of Datamarts - Data technical referent - Cloud Management (Google Cloud Platform)

- Developed an AI system for Dog's Activity Recognition. - Concurrently spearheaded the research and development division to investigate alternative deep learning methods. - Collaborated with a Machine Learning Engineer to handle project-related data. - Executed an end-to-end machine learning project. - Established tools to enhance project management, data visualization, and workflow. - Ensured all implemented tools were open source for transparency and collaboration.

- Implemented a product-based recommender system to enhance CRM development. - Utilized Association Rules, interest scoring, and segmentation techniques in CRM development. - Deployed time series forecasting using RNN to predict call center activity and effectively scale teams. - Contributed to the development of a Speech-to-Text Proof of Concept using the WaveNet algorithm from DeepMind. - Applied NLP algorithms to examine and analyze survey data.

- Implementation of the Data Management Platform (DMP) - Data Muggling with the platform Cloudera (Hive, Spark...) - Implementation of ML's algorithm -Build Datalake

- Reporting to monitor the activity - Production of KPI to pilot the activity - Setting up the calculation of bonuses for employees