TikTok

• ML lead in iOS14 ad recommendation calibration. 
• ML research in data enrichment, incl. NLP tag extraction and GNN embeddings.
• ML modeling on ads recommender models in the Brand Ads team in various areas such as position bias modeling, model optimization and multi task learning.

Senior Machine Learning Engineer

ChelleStudio

ChelleStudio empowers entrepreneurs, helping them to develop and grow their app. Our portfolio of apps currently registers 300k+ new users a month. Our team develops state-of-the-art AI/ML tools to support the growth and development of our apps.

Founder & CEO

Shake (Acquired)

Acquired in Sept. 2023

Leading a team of 4 machine learning scientists and 1 mobile developer to work on the development and growth of Shake. Collaborated closely with stakeholders such as investors and publishers.
• Registered 100k+ users
• Built a community of 100+ dance teachers and content creators
• Developed an on-device and in-realtime AI pose estimation

Peeps

Peeps is a friend discovery app that attracts over 500k monthly active users. Featured in the top 20 US Social Networks.

Acquia

• Machine Learning and Data Science within the Data Science Team of AgilOne-Acquia. 
• Worked on distributed deep learning models, based on Horovod for Tensorflow and Spark, with custom Gradient Descent on multiple Spark executors. 
• Applied research on deep learning models. Implemented a model for Price and Discount Optimization on Keras and Tensorflow.

Machine Learning Engineer

• Data analysis, machine learning prototyping and test-driven development of the core machine learning platform.
• Used collaborative filtering with distributed ALS.
• Prototyped and implemented a distributed recommender system analyzing more than 60 billions events each day.
• Used Pandas, Spark, PyTest, SQL, Python and a custom Machine Learning framework.

Data Scientist @AgilOne (acquired by Acquia)

Infosys

OCR and NLP solutions to cluster and extract data from PDF invoices.
• Used neural networks, KNN Neighbors and Naive Bayne classifiers to improve the data extraction coverage from 10% ( old system ) to 80%.  Reached an accuracy of 78%.
• Created a new clustering model for data extraction from documents to make the system independant from the layout of the document.
• Helped deploying the model in production, handling 1000+ documents a day. 
• Used NLP, Tensorflow, Keras and Python.

AIROD Technologies

• Research and Prototyping on terrestrial and marine robots.
• Programmed on Robot Operating System ( ROS ) on a Raspberry.
• Developed an IoT project with SigFox.

Robotics and Software Engineer

SmartX

Development of a model for autonomous swarm flight of drones. 
• Created a model based on genetic neural networks to simulate the flight of drones ( no communication between drones, exterior information form the sensors only). 
• Applied Boid's theory to improve the neural network. 
• Used Java, OpenCV, Arduino, Genetic Neural Networks.

Research in Machine Learning and Robotics

Sold my first AI startup at 27, empowering others to do the same. Passionate about building cool products.

Taplio

Valentin Chelle's Linkedin Analytics

Want to drive more opportunities from LinkedIn?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Valentin Chelle's Linkedin Analytics

Want to drive more opportunities from LinkedIn?

Valentin Chelle's Linkedin Analytics