Dmitry Kovalev Email & Phone Number

Netherlands

Amsterdam, North Holland, Netherlands

Workera.ai is the leading AI-powered skills intelligence platform that enables organizations and individuals to precisely measure and develop technical capabilities. Using advanced psychometrics and machine learning, Workera has assessed 3.2M+ skills across 24 countries, creating 75k+ personalized learning paths. Founded by Stanford lecturer Kian Katanforoosh and chaired by AI pioneer Andrew Ng, Workera is revolutionizing how companies understand and develop their technical talent.Key Responsibilities & Technical Expertise:* Developed custom Deep Knowledge Tracing models using transformer architectures and text embedding techniques* Optimized experimentation workflows, model training pipelines, and performance tracking systems* Integrated Large Language Models (LLMs) into existing applications, including prompt engineering* Designed and implemented scalable multi-agent environments for automated skill assessment* Enhanced performance tracking systems using modern ML observability tools

Amsterdam, North Holland, Netherlands

Innovative medical technology company specializing in AI-powered solutions for rapid stroke diagnosis and treatment.- Developed computer vision solutions for medical devices used in stroke treatment- Trained and iterated on machine learning models to improve diagnostic accuracy- Built backend services on Google Cloud for secure collection and anonymization of medical data- Collaborated with medical specialists to research and implement innovative approaches in stroke diagnostics and treatment

Warsaw, Mazowieckie, Poland

Innovative ML-driven weather prediction startup built from scratch. Founded by an experienced leader who previously built and successfully sold multiple AI/ML companies to tech giants including Google and Pinterest.• Pioneered cutting-edge weather prediction systems using aerial and radar spatial data, leveraging diverse deep learning architectures: Advanced Autoencoders and ConvLSTM networks, transformer-based architectures and VAEs and GANs for weather pattern synthesis.• Built and automated ML pipelines in a cloud-based environment (AWS Sagemake);• Integrated models into the production environment, implementing scalable inference from various sources (RabbitMQ/Celery), and optimizing models and backend performance (ONNX);• Led a small team of ML engineers, responsible for task planning and prioritization, and onboarding.

Minsk, Belarus

Lyft (NASDAQ: LYFT) is a leading transportation technology company and one of the largest ridesharing platforms in North America. With a market cap in the billions and serving millions of rides daily, Lyft continues to innovate in mapping technology and transportation infrastructure. The company leverages advanced machine learning and computer vision to enhance its mapping and routing capabilities across its nationwide network.• Designed and implemented ML pipelines for automated routing map enrichment utilizing PyTorch, Kubernetes, Kubeflow/LyftLearn/Flyte, PySpark, AWS Services, and Weights&Biases;• Conducted experimentation with cutting-edge computer vision architectures for object detection and segmentation (EfficientDet, MobileNet, U-Net, YOLO);• Optimized and deployed production models to mobile devices using TensorFlow Lite and ONNX;• Managed and processed large datasets, ensuring version control and accurate collection. Developed an automatic image quality assessment service utilizing PyTorch and OpenCV;• Facilitated cross-team collaboration and integrated ML tools into back-end and mobile services.

Minsk, Belarus

Juno (acquired by Gett for $200M) was a NYC-focused ride-hailing platform that pioneered a driver-first approach, offering lower commissions and innovative equity programs. The company competed directly with major players in the ride-sharing market before ceasing operations in 2019.• Developing machine learning and data processing services using: Python, Pytorch/TensorFlow, Docker, Kinesis, Prometheus, PySpark, Luigi.• Research on solving the following tasks: fraud detection, forecasting supply/demand, CLV(customer lifetime value): XGBoost/Catboost, TensorFlow, scikit-learn, pandas/numpy, PySpark/Zeppelin.• Getting insights from big amounts of data and preparing reports for business needs. For example, analyzing bottlenecks in demand/supply activity for the ride-sharing market: Python, PostrgeSQL, Jupyter, Amazon S3, Tableau.• Working with A/B tests: collecting data results analysis.

Minsk, Belarus

Vention is a global software development powerhouse with 20+ years of experience, serving 500+ clients from startups to Fortune 500s. The company has supported 20+ client IPOs and contributed to $15B+ in client acquisitions, with an average client engagement of 36 months. Partnering with world's top tech leaders, Vention delivers comprehensive software solutions across 30+ verticals and 25+ technologies.- Developing a web application for financial analysts (WisdomTree) based on Sitecore CMS (got a Sitecore certificate for this project). Areas of responsibility: backed development (C#, ASP.NET MVC and REST API), searching (Apache Solr) and fronted (JavaScript). Based on Microsoft Azure environment- Improving aae.org application(healthcare): C#, JavaScript, HTML5/CSS3- Hosted communication process with customers: meetings, reports, demos;- Newcomers adaptation activity

Minsk, Belarus

EPAM Systems (NYSE: EPAM) is a global leader in digital platform engineering and software development services, serving Fortune 500 clients worldwide. With 20+ years of experience and recognized by Forbes, Fortune, and Forrester as a top technology service provider, EPAM delivers enterprise-scale digital transformation solutions.- Development and supporting of Enterprise Resource Planning (ERP) system for managing manufacturer processes of USA major industrial companies;- Performance debugging and improvements of .NET Cloud application (SQL and ASP.NET Web Services optimizations);- Feature demos and cross-team meetings;