Vitalii Lykhkun

Led a 7-engineer Gemini applied-AI workstream delivering production LLM capabilities for Google Cloud enterprise customers. Directed the launch of a Gemini-powered enterprise search capability that generated $28M in incremental ARR during its first year.

Owned ranking and recommendation improvements for Facebook News Feed candidate generation, spanning model development, distributed training, experimentation, and privacy review. Designed a two-tower PyTorch retrieval model that increased candidate recall by 12.4% and produced a 3.1% sitewide lift in time spent on Feed during A/B testing.

Architected a 3-layer enterprise platform using React and Next.js, Node.js services, PostgreSQL, and AWS, supporting secure financial and administrative operations. Engineered REST APIs and modular backend services across authentication, authorization, data processing, and third-party integrations.

Modernized product discovery, inventory visibility, and personalized shopping through React and Next.js experiences backed by Node.js, PostgreSQL, AWS, and Docker. Designed API-first services across catalog, customer profiles, search, recommendations, and order processing.

Built responsive customer-facing interfaces and reusable components across JavaScript, HTML, and CSS. Resolved application defects through reproduction, root-cause analysis, and verified remediation in partnership with senior engineers and QA.

Bachelor Degree in Computer Science from Lviv Polytechnic National University from 2000 to 2005.