Juice is the first generative audio platform, designed to deliver existing audio content (music, podcasts, etc.) or generate it in real-time for each user (news, practical information, B2B content).Key Achievements: • Designed and launched a consumer-facing product with over 20,000 users. • Conducted rapid iterations to meet user expectations, improving App Store rating from 4.3 to 4.6 based on 100+ reviews. • Built a B2B model to address a user problem, signing three major accounts including MACIF, accompanied by a joint press release.Technical Accomplishments: • Monitoring Engine: Developed an engine that continuously listens to thousands of sources (classic media, social media), filtering and sorting them using AI/NLP algorithms. From R&D to MLOps. • Jungle, Automated Audio Content Generator: Takes into account user context and data as well as external data like APIs to engage users. • Recommendation Algorithm: Created a system to recommend the most relevant content based on context and user preferences, adaptable over time (collaborative filtering). • Text-to-Speech: Developed our own realistic synthetic voices using state-of-the-art deep learning techniques. • Speech Recognition: Developed an in-house, low-level (C) speech recognition model embedded in smartphones for maximum efficiency. Implemented a dynamic AI model update pipeline that doesn't require app updates. • Data Privacy & Efficiency: Developed an in-house Kafka/SQL pipeline for ultra-efficient, anonymized tracking of listening data to optimize content ratings.

NLP Research and Development• Conducted research and development in Natural Language Processing (NLP) for a software company serving large enterprise clients in industries like banking and insurance.• Deployed APIs, NLP tools, and chatbots for client services.• Engaged in technological surveillance in the fields of Deep Learning and NLP in Silicon Valley. Participated in major regional gatherings such as conferences, meetups, and competitions including events in San Francisco, Stanford, Baylearn, NeurIPS, and Nvidia GTC.

• Conducted research in the Natural Language Processing (NLP) laboratory at the University of Michigan (Language and Information Technology - LIT) under the direction of Rada Mihalcea.• Collaborated with the Medical School to create a dataset and establish baselines for an NLP task.• Authored a scientific paper under the guidance of PhD candidate Mahmoud Azab, which was accepted at EMNLP 2019 https://aclanthology.org/D19-1122.pdf

I was a software developper at the Applied Innovation Exchange of Capgemini Suresnes.- Involved in prototyping and beta testing on robotic platforms, including drones and humanoid robots like Nao and Pepper.- Conducted rapid iterations with clients from large retail and transport sectors.- Client demonstrations, simplification, and training in digital transformation for major companies.

• Prepared and organized our school's TEDx event, leading a team of 10 people• Sold out event with over 150 participants• Hosted 6 inspiring speakers, including the co-founder of Ornikar

Summer analyst