Temirlan Zarymkanov

Software Engineering Team Lead

Adigy.ai

amazon ads management

Leads software engineering and infrastructure; reports 20% higher user revenue, 35% lower monthly costs, and reduced downtime.

• Leading team of software engineers, responsible for architecture design, cloud infrastructure management, security and network support. • Implemented smart state management system for bid optimization with weighted average for different markets, increasing user revenue by 20%. • Reworked infrastructure, optimized database queries, and collaborated with external service providers (AWS, datadog), reducing monthly cost by 35%. • Migrated several services to better, cheaper alternatives, boosting monitoring and observability.

ML & Software Engineer

Sewts GmbH

robotics and automation

Built and deployed ML pipelines and synthetic textile data generation; reports 10% accuracy improvement and 90% lower labeling requirements and costs.

• Designed and maintained robust, reproducible, containerized ML pipelines (PyTorch + Docker) for both batch and real-time inference, enabling continuous retraining on live customer data and improving accuracy by 10%. • Integrated active learning strategies into the data annotation loop, cutting manual labeling requirements and associated costs by 90%. • Deployed end-to-end ML models (PyTorch Lightning) and supported full-stack infrastructure (TypeScript, Python).

Software Engineer / Team Lead

Fleek.kz

logistics and optimization platform

Led development of a logistics platform and delivered its first MVP in two months, twice as fast as planned.

• Led development of an logistics platform using FastAPI and Next.js; implemented route optimization models and automated shipment tracking. • Designed scalable architecture and implemented modular, maintainable backend and frontend systems based on business requirements, ensuring rapid iteration and feature delivery. • Deployed the first MVP in just 2 months — 2x faster than the original timeline — enabling early customer feedback and faster market validation. • Acted as Scrum Master: facilitated sprints, daily standups, and sprint planning to keep a fully remote team aligned and productive.

Software Engineer

Macdent.kz

Worked on full-stack development and infrastructure, reporting faster response and load times, improved uptime, and doubled team development speed.

• Full stack software development with PHP, JavaScript, decreased bounce rate by 15% on key pages due to improved feedback during data loads. • Created automatic backup system for database, added sharding, and improved SQL queries to decrease database answer time by 1000%. • Optimized database queries and introduced predictive caching, improving system response time by 10x. • Deployed server on Cloud (AWS), increasing load speed by 5 times, ensuring 99p uptime. • Set up CI/CD and increased development speed in the team by x2.

Machine Learning Researcher

Nazarbayev University

Research university in Kazakhstan.

Researched applied ML and computer vision, published two peer-reviewed papers, and improved a mobile pest-detection model by 50%.

• Published two peer-reviewed papers in applied machine learning and computer vision using PyTorch and Sklearn. • Developed retraining pipeline for mobile pest detection system, improving model performance by 50%. • Enhanced model interpretability and explainability while maintaining high accuracy. • Optimized YOLOv5 architectures on CUDA for faster inference on edge devices.

Master of Science, Data Analytics (Machine Learning, Computer Vision, and Optimization)

University of Hildesheim

Bachelor of Science, Computer Science

Nazarbayev University