Vladimir Lazyuk

Lead AI Python Engineer and Team Lead overseeing production-grade AI systems powered by LLMs and multi-agent architectures. Implemented a YOLO-based defect detection system for conveyor lines with real-time video streams from shop-floor cameras to identify safety violations, process deviations, tardiness, and downtime. Deployed automated product rejection on production lines. Built a call analytics platform integrated with CRM (Bitrix24, AmoCRM) and A1 telephony, including transcription, call classification, and multi-agent quality scoring scripts with analytics dashboards. Led creation of an enterprise video conferencing platform with LLM-based meeting minutes, with on-prem deployment and integration with Google Meet and Discord. Developed LLM-based document processing with voice control and multi-agent workflows. Managed infrastructure architecture (70% on-premise, 30% cloud), Kubernetes, staging and production environments; self-hosted GitLab, Jira, and Drive; led a cross-functional

Led LLM and automation projects for Russian-speaking international companies, delivering end-to-end AI pipelines and automation solutions for diverse domains, including document processing, data extraction, and integration with enterprise systems.