Hire Dmytro I. - Data Scientist

Owned data infrastructure for Snapchat's AI chatbot, and built it from scratch: storage and pipelines for chat data under strict privacy, legal, and security requirements. That infrastructure is still the org's source of truth for engagement, quality, and safety reporting. - Turned a classification approach built for one product into a company-wide framework spanning text, image, and video, then got it approved by finance and legal for reuse. Teams across search quality, ads relevance, and maps evaluation now build on it instead of starting from scratch. - Built and ran the quality and safety evaluation process for the product, including a human-labeling review workflow and a validation standard now used to check ML model outputs against a shared benchmark dataset. -Weighed vendor LLM cost and control against flexibility, then moved a keyword-expansion pipeline to self-deployed open-source models. That cut cost and let me rebuild the tool to support multiple LLM providers.

* Designed and delivered a full Product Analytics course at Projector tech institute — taught across three cohorts over multiple months each, combining lectures with hands-on practical sessions. * Built the entire curriculum from scratch, covering the full product analytics toolkit: metric design, A/B testing and experimentation, funnel and cohort analysis, and data-driven product decision-making * Delivered both lecture and practicum formats, grounding theory in real-world case studies from production AI/ML systems. * Graduated 50+ students across three course runs * Ran the course as a charity initiative — 100% of tuition proceeds were donated to charitable causes

* Advised more than 10 startups and companies on product analytics, ensuring the right tools were in place. * Guided the hiring process and team building, to help them build own Data Science team from scratch * Developed OKRs and KPIs, establishing measurement systems to track progress and support business goals.