Samuel Prevost
Verified Expert in Engineering
Artificial Intelligence Developer
Samuel is an artificial intelligence engineer with expertise in building and deploying deep learning models. He is well-versed both in the scientific literature and the reality of developing complex systems, as well as communicating on technical challenges and their state-of-the-art solutions. Samuel has directed R&D projects in partnership with universities and contributed to open source technologies.
Portfolio
Experience
Availability
Preferred Environment
Linux, Vim Text Editor, Oh My Zsh, SSH, JetBrains
The most amazing...
...edge computing model I've developed is used by Mercedes-Benz to help their car dealers check returning vehicles by detecting damages as they drive by.
Work Experience
Data Scientist
FaceCake Marketing Technologies
- Implemented a physic-based rendering 3D engine into a high-performance media processing pipeline for real-time mobile rendering and pose estimation.
- Maintained a Google project's private fork and improved its build system to suit our needs.
- Implemented a multiplatform CI/CD pipeline between repositories in order to partition the work environment of different teams and increase their productivity.
- Built an Android demo application with functional programming-oriented Kotlin that uses our self-contained SDK to showcase 3D models of products in real time, using 3D pose estimation.
Lead Data Scientist
Atos
- Developed a serverless-based social network crawler that gathered 70,000 posts (with images) from hundreds of public pages in under 24 hours and fulfilled customer requests asynchronously.
- Built a transfer learning-based low dimension embedding network to drive social network marketing campaigns of large companies.
- Designed and built interactive web interfaces with Plotly's Dash to visualize, compare, and predict the impact of marketing campaigns in different countries.
- Maintained a large, private codebase, using version tracking tools, setting up CI/CD pipelines, and training team members to write seamless uniform code with documentation available through a private wiki, using Sphinx.
Chief AI Officier
TwoTronic
- Developed a computer vision model, running on an embedded system, that detects and classifies defects on cars from pictures taken as they drive by.
- Built a web interface to demo the model using Plotly's Dash framework, which made a gRPC call to a TensorFlow model hosted on an on-premise TensorFlow Serving server.
- Oversaw the rollout and update of models on many edge devices, using A/B and canary testing and profiling performance.
- Directed a research team that competed against two other national research institutes to deliver the best prototype model in the least amount of time.
Research Manager
Sorbonne University
- Communicated with interdisciplinary researchers about the technical capabilities of the state of the art in artificial intelligence.
- Researched new style transfer techniques and their explainability and implemented those techniques in PyTorch through a neural ordinary differential equation (ODE)-like network architecture.
- Built a comprehensive bibliography of the topic and attended the 33rd Annual Conference on Neural Information Processing Systems (NeurIPS | 2019) to discuss research prospects with world-leading authors.
Experience
High-level Embedding on Social Media Content
AI in Art History: Delacroix Project
http://www.correspondance-delacroix.fr/We presented our work on style transfer at the NeurIPS | 2019 convention, and the project is now pending approval for a grant from the European Research Council. This project would revolutionize the field of art history by bringing new analysis techniques, never before used in the domain, and it could push the state of the art in style transfer and deep learning explainability.
Hide an Image Into Music
https://github.com/sam1902/SpectroGenV2Learn and Recognize
https://github.com/sam1902/Learn-n-Recognize-CppNewton Multivariate Solver
https://github.com/sam1902/NewtonMethodMultivarSimplex Algorithm Solver
https://github.com/sam1902/SimplexSkills
Languages
Python, Bash Script, SQL, C++, Kotlin, Java
Libraries/APIs
Keras, NumPy, Scikit-learn, PyTorch, TensorFlow, OpenCV, Pandas, Protobuf, OpenGL ES, SpaCy, Matplotlib, Eigen
Tools
Git, JetBrains, Vim Text Editor, Oh My Zsh, Bazel, CMake, Plotly, AWS CloudFormation, Asana, Audacity, LaTeX
Platforms
Linux, Amazon Web Services (AWS), Docker, Android, NVIDIA CUDA
Other
Deep Learning, Computer Vision, R&D, SSH, Learning Transfer, Convolutional Neural Networks (CNN), Zotero, Artificial Intelligence (AI), Linguistics, Relational Algebra, Entity-relationships Model (ERM), Machine Learning Operations (MLOps), Data Cleaning, Fourier Analysis, Software Architecture, Cryptography, Data Structures, Dash, TensorFlow Serving, LabelMe, Binary Classification, Mask R-CNN, 3D Rendering, 3D Pose Estimation, Sentiment Analysis, Natural Language Processing (NLP), Spectroscopy, Tech Conferences, Machine Learning, Edge Computing, Real-time Embedded Systems, Optimization, Calculus, GPT, Generative Pre-trained Transformers (GPT)
Frameworks
Flask, gRPC
Paradigms
Functional Programming, Linear Programming, Dynamic Programming, Serverless Architecture, REST
Storage
Amazon S3 (AWS S3), PostgreSQL, NoSQL, SQLite
Education
Bachelor's Degree Completion (Exchange Program) in Computer Science
University of Montreal - Montreal, Canada
Bachelor's Degree in Computer Science
Sorbonne University - Paris, France
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
Share your needs
Choose your talent
Start your risk-free talent trial
Top talent is in high demand.
Start hiring