Artsiom Yancheuski
Verified Expert in Engineering
Software Developer
Artsiom is a highly dedicated software engineer striving for technical excellence. Current interests are mostly in distributed systems, AI pipelines, asynchronous workflows, and general CS algorithms. Artsiom excels at managing and leading teams and, at the moment, prefers hands-on development and technical expertise.
Portfolio
Experience
Availability
Preferred Environment
Artificial Intelligence (AI), Pipelines, Docker, Google Cloud, Python
The most amazing...
...system I've designed and built involved highly scalable AI computer-vision pipelines that powered mission-critical processes in one of the world’s top startups.
Work Experience
Lead Python Engineer
Revolut
- Designed and created a distributed system that processed mission-critical data via ML based on an event-driven architecture with mediator topology. I wrote most of the pipeline myself, along with supervising the production launch and maintaining it.
- Redesigned a Python API to improve its efficiency and maintainability via restructuring, caching, improved logging, and a separate server with separate infrastructure and a CI/CD pipeline.
- Rewrote a vast legacy codebase from 50,000 LOC into 15,000 LOC, separated it into single responsibility repositories with dedicated CI/CD pipelines, and launched it in production.
Lead Software Engineer
EPAM Systems
- Designed and developed end-to-end numerous pipelines to intelligent automation and artificial intelligence for Fortune 500 companies.
- Drove intelligent automation and artificial intelligence COE deployment and operations to maximize the return on in-house automation.
- Created (from the ground up) solutions for the processing of digital workloads from source input through OCR and extraction all the way into the model improvement pipelines and quality control.
- Used WorkFusion, Abby SDK, and Groovy to deploy pioneering artificial intelligence solutions.
- Supervised several projects at the same time as a team lead and delivery manager.
Experience
Distributed System for Machine Learning Workflows
I served as the architect and main implementation engineer and was there from inception to release and maintenance.
The system allows for the asynchronous submission of input for the processing and results polling on the interface part. Internally it uses Celery encapsulating Redis for the nonblocking parallel processing of large amounts of data.
The tasks are managed in mediator topology, with a separate workflow layer responsible for cross-cutting concerns and task dispatching. We deployed the system in a Kubernetes cluster on the GCP. The system includes numerous AI models produced by multiple data scientists in parallel that are deployed as separate docker images.
The results are aggregated and served back to request with high availability.
Skills
Languages
Python 3, SQL, Python, CSS, HTML, Java, Groovy, JavaScript
Frameworks
Flask, Django
Tools
Celery, Apache Airflow, Pytest
Libraries/APIs
Pandas, Scikit-learn, SQLAlchemy
Platforms
Kubernetes, Docker, Google Cloud Platform (GCP)
Storage
Google Cloud, PostgreSQL, Exasol
Other
Pipelines, Artificial Intelligence (AI)
Education
Bachelor's Degree in Finance
Belarusian State University - Minsk, Belarus
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