Dmitry Mezhensky

Dmitry Mezhensky

Dmitry Mezhensky joined Grid Dynamics in 2014 and has worked on various Big Data projects since. One of the major projects, iCrossing, was a huge success as we built a high-performing Big Data platform. Dmitry is currently on-site at a large retailer.

Blog posts:

Dmitry Mezhensky

Dmitry Mezhensky joined Grid Dynamics in 2014 and has worked on various Big Data projects since. One of the major projects, iCrossing, was a huge success as we built a high-performing Big Data platform. Dmitry is currently on-site at a large retailer.

Dmitry Mezhensky

A scalable, configuration-driven machine learning platform

Co-created by Grid Dynamics Director of Data Engineering, Dmitry Mezhensky, and Yieldmo Head of Analytics and Data Science, Sergei Izrailev Introduction Yieldmo, a Grid Dynamics client, is an advertising platform that helps brands improve digital ad experiences through creative tech and artificial intelligence (AI). The company uses bespoke ad formats, proprietary attention signals, predictive format

LLMOps blueprint for closed-source large language models

Building solutions using closed-source large language models (LLMs), including models like GPT-4 from OpenAI, or PaLM2 from Google, is a markedly different process to creating private machine learning (ML) models, so traditional MLOps playbooks and best practices might appear irrelevant when applied to LLM-centric projects. And indeed, many companies currently approach LLM projects as greenfield

Analytics and ML platform modernization: A case study

MLOps and DataOps principles, such as infrastructure-as-a-code management, continuous integration and continuous delivery, proper monitoring, and a standard approach to working with data assets, are essential components of a modern data estate. In this case study, we show how we helped a global gaming loyalty company improve business KPIs, such as reduced total cost of

How to enhance MLOps with ML observability features: A guide for AWS users

Adoption of machine learning (ML) methods across all industries has drastically increased over the last few years. Starting from a handful of ML models, companies now find themselves supporting hundreds of models in production. Operating these models requires the development of comprehensive capabilities for batch and real-time serving, data management, uptime, scalability and many other

Enterprise-grade ML platform in AWS: A starter kit

Grid Dynamics has developed an ML Platform Starter Kit for AWS that provides a reference architecture and capabilities for building an ML platform, including an experimentation pipeline, automated CI pipeline, and continuous deployment and serving infrastructure management. The platform utilizes AWS SageMaker and MLflow for model development, training, and serving, and includes CI/CD automation and a model serving layer built on top of Kubernetes and SageMaker.

Data quality control framework for enterprise data lakes

Data quality control is a critical capability for businesses, as data quality issues can disrupt processes, invalidate analytics, and damage a company’s reputation. However, data quality control is often undervalued, and there is room for improvement in most enterprises. Grid Dynamics has developed a scalable and easy-to-extend data quality control framework that integrates with various data sources, performs data quality checks, and visualizes the results using open-source tools like Soda SQL, Kibana, and Apache Airflow.