Florida Atlantic University

South Ural State University

CTO at Ople

Skills: Software Development · Artificial Intelligence (AI) · Engineering Management

Branch

Senior Principal Data Scientist

Defined and executed the AI and ML strategy for the company, identifying key technologies and engineering infrastructure required to drive success.
Collaborated closely with leaders across the organization to ensure the delivery of high-quality customer solutions, aligning with their needs and expectations.
Provided comprehensive roadmaps for multiple critical projects, guiding the team's execution with Agile methodologies.
Developed the initial version of the model serving software specifically designed for industrial edge devices, enabling efficient and reliable deployment of machine learning models.
Lead the development and architecture of an internal MLOps cloud platform, facilitating the management of deep learning model lifecycles on industrial edge devices, including seamless over-the-air model updates.
Drove the end-to-end development of several deep learning models applied to X-ray images, taking them from inception to production for several customers.
Conducted extensive research and development in computer vision hardware, encompassing cameras, lighting, lenses, and mechanical design for industrial equipment.
Successfully organized and executed multiple proof-of-concept (POC) projects with potential customers, showcasing the capabilities and value of the company's AI and ML solutions.

Greyscale AI

Vice President of AI and ML

Successfully sourced and recruited highly qualified staff members to establish and strengthen teams, scaling the engineering team from 0 to 20 engineers. Built a diverse team of full stack developers, ML engineers, data scientists, and UI/UX designers.
Envisioned, developed, and executed a comprehensive engineering roadmap, providing clarity to engineers regarding the company's strategic objectives and the bigger picture.
Spearheaded end-to-end management of the overall system architecture while actively contributing to the development of critical backend components.
Developed the ML backend for a stealth product focused on semi-supervised learning, demonstrating expertise in driving customer support and deployments.
Led research and development (R&D) efforts, resulting in a 50% increase in platform efficiency. Governed research endeavors on proprietary AtoML technology, enhancing the capabilities of Ople's products.
Provided technical leadership to the team, guiding the successful delivery of Ople's products from ideation to market launch.
Led architectural changes that effectively reduced the AWS bill by 25%, optimizing resource allocation and cost-efficiency.
Built and led a team of engineers responsible for developing entire products for renowned companies such as Dupont and Zappos.
Organized and executed multiple proof-of-concept (POC) projects with potential customers, showcasing the value and capabilities of Ople's platform.
Actively engaged with enterprise customers, ensuring smooth onboarding onto the Ople platform and fostering strong ongoing relationships.

Ople

Developed several POCs for the future Ople’s products.
Developed core ML backend functionality using Python,  Flask/Celery, xgboost, scikit-learn, Pytorch, pandas. 
R&D in the area of AutoML and hyperparameter optimization.
Responsible for architectural design of the Ople platform POC.

Founding Engineer

Launched multiple innovative improvements and received the Sentient Innovator Award and a patent. 
Utilized existing algorithms and applied them to domain-specific applications, contributing to the development of the Sentient interactive shopping experience found on Shoes.com, Shoeme.ca, and Zappos.com.
Created a tool on top of Torch7, enabling training of deep convolutional models for product similarity using triplet learning techniques.
Played a key role in building a distributed inference engine to serve Torch models efficiently and effectively.
Defined approaches for building deep learning models specifically focused on content embedding, leading to the creation of patents and empowering the development of real-world products.

Developed Linear Algebra, hyperparameter optimization, image segmentation and dimensionality reduction (t-SNE) modules for Sentient’s deep learning platform in Java and C.

Sentient Technologies, Inc.

Senior Software Engineer

Presentation and talks
1. P. Tsatsin, V. Beskachko “The influence of the magnetic field on the inertial
deposition of the particles on the surface of the rotating disk”, 13th International
Conference on Liquid and Amorphous Metals, Ekaterinburg, Russia, 2007
2. P. Tsatsin, W. Tichy “Predicting the final spin of the black hole mergers”, APS
2010 April Meeting, Washington, DC
3. P. Tsatsin (FAU), Reuben Budiardja (ORNL), Christian Cardall(ORNL), Eirik
Endeve (ORNL), Cory Hauck(ORNL), Pedro Marronetti(FAU), Bronson
Messer(ORNL) and Anthony Mezzacappa(ORNL) “ GenASiS: A full GR-RMHD
simulation framework: overview, goals, and preliminary tests”, APS 2011 April
Meeting, Anaheim, CA
4. P. Tsatsin (FAU), Reuben Budiardja (ORNL), Christian Cardall(ORNL), Eirik
Endeve (ORNL), Cory Hauck(ORNL), Pedro Marronetti(FAU), Bronson
Messer(ORNL) and Anthony Mezzacappa(ORNL) “ GenASiS: A full GR-RMHD
simulation framework: overview, goals, and preliminary tests”, 6th
Annual Gulf
Coast Gravity Meeting, Florida Atlantic University Boca Raton, May 2011
5. P. Tsatsin, P. Marronetti and K. Yakunin “Simulations of the neutron star binaries
with an arbitrary spins using Einstein Toolkit”, APS 2012 April Meeting, Atlanta,
GA

Graduate Student and Teaching Assistant

As a Graduate Student at Florida Atlantic University I've been involved in research projects in Computational Astrophysics and Numerical Relativity. This allowed me to gain a wide experience in the following areas: 
- problem solving and conducting individual and team research projects (work in collaboration with Oak Ridge National Laboratory and NINJA project https://www.ninja-project.org ) 
- high-performance scientific computing and physical simulations  
- Mathematics and numerical algorithms especially for ODEs and PDEs 
- developing scientific computational codes (C/Fortran) 
- Linux, bash programming 
- working experience with supercomputers: Jaguar, Smoky, Kraken, Quenn-bee, Lens, Stampede 
- visualisation experience with Visit, awk+gnuplot 
- presentig of the scientific results on the meetings 
- 2 publications

Graduate Student

I've got my first research experience when I was working on my Bachelors Thesis and lately on my Master Thesis at South Ural State University. I've been involved in the project in area of the Physicochemical Hydrodynamics. Particularly, we were investigating the theory introduced by Veniamin Levich of inertial deposition of the microparticles suspended in the  conducting fluid flow  under rotating disc (magnetohydrodynamics (MHD)) analog on classical von Karman swirling flow problem) and also influence of the external magnetic field on such an effect. The symmetry of the problem allows to reduce full set of MHD equations to a two point boundary value problem for system of the non-linear ODEs, which was solved utilising Multi-shooting method. The equations of motion of the micro-particle can be easily  integrated given the known fluid velocity field.  All these numerical methods I implemented in the code which also has GUI and subsequent graphical visualisation of the   particles trajectory.

Petr Tsatsin