Aliaksandr Tsukanau

Aliaksandr is a machine learning engineer at Vention with hands-on expertise in computer vision, large language models (LLMs), and time series analysis. At Vention, he’s played a key role in building core AI systems for fast-scaling startups and enterprise clients, solving complex challenges across constrained environments and high-scale deployments.

He implemented several critical infrastructure initiatives that allowed the LLM inference platform to scale to more than four billion tokens served daily via OpenRouter, and led an AI-driven operating room efficiency solution designed to operate with limited data and compute resources. His experience spans medtech, retail, banking, and adtech, where he has streamlined data workflows and developed ML-powered systems that enhance operations and decision-making.

Aliaksandr specializes in Python-based machine learning systems with a strong focus on MLOps tooling, deployment automation, and runtime optimization. He builds scalable pipelines for model training, inference, and monitoring, and continually improves resource usage and latency across production environments.

• LLM inference at scale (vLLM, sglang, Ray Serve, token throughput optimization)
• Computer vision and time series ML applications
• AI solutions under data and compute constraints
• MLOps and deployment automation for ML pipelines
• Scalable infrastructure and model performance tuning
• Cross-domain ML engineering (medtech, retail, banking, adtech)
• Backend Python development
• ML integration into production systems