Staff Software Engineer

Zhachory
Volker

Building intelligent systems at scale — from search ranking at YouTube to ML infrastructure at Rokt.

View Resume Get in Touch
GitHub LinkedIn me@zhachory.com

About

I've spent 10+ years building ML systems, search ranking, and recommendation engines at YouTube, Google, and Rokt. The systems I've worked on serve billions of queries and the infrastructure choices I've made have saved millions of dollars.

My focus is the intersection of ML engineering and product impact — turning research into production at scale. I care about responsible AI, computational journalism, and the systems that shape how people find and trust information.

I'm looking for staff or principal engineer roles leading technical strategy on AI/ML-intensive systems. Based in New York City.

Zhachory Volker

Experience

PDF Resume ↗
  1. Oct 2025 — Present Rokt

    Staff Software Engineer

    Leading and mentoring a team of 15 engineers on the Audience and Bidding team. Architected and deployed a dynamic ranking layer for on-the-fly audience generation, scaling revenue from generated audiences by 120%. Reduced ML audience generation pipeline costs by 75% (~$750k annually) via migration to a two-tower infrastructure with a robust data refresh system. Spearheaded an org-wide AI integration initiative improving Developer Satisfaction by 25%.

  2. Mar 2018 — Sep 2025 YouTube / Google

    Senior Software Engineer, YouTube Search

    Designed a unified ranking framework boosting Search Active Users by 2% and CTR by 25% across multiple features in a single launch. Led the YouTube Search News experience with a team of 4, increasing News CTR by 38% and daily News Searches by 10% (~8M queries) through clustering algorithms and NLP models. Led a resource optimization project achieving 60% compute reduction (estimated $1.8M annually).

  3. Mar 2017 — Mar 2018 Google Research

    Software Engineering Resident

    Applied NLP and ML (RNNs, CNNs, Decision Trees) to 10,000+ articles for Google's fact-checking corpus, results still surfaced on Google today. Built MapReduce pipelines in C++ utilized by six major news publishers. Developed label propagation and semi-supervised learning pipelines on large-scale graphs, increasing library adoption by 8%.

  4. May 2015 — Mar 2017 Ericsson

    System Data & Performance Intern

    Full-stack development in Polymer, Python, and Java. Data mining pipelines contributing to a 12% reduction in site risk incidents.

Education

2013 — 2018 University of North Texas

B.S. Computer Science — Dean's List

Projects

Skills

Languages
Python, C++, Go, Java, JavaScript
ML & AI
PyTorch, TensorFlow, Keras, LLMs, NLP, Ranking Systems, Recommender Systems, Two-Tower Architecture
Data
SQL, BigQuery, Spark, MapReduce, ETL, Embeddings
Infrastructure
GCP, AWS, Kubernetes, Docker, Terraform, Borg
Search
Information Retrieval, Query Understanding, Embedding Search, Indexing

Get in Touch

Open to staff and principal engineer opportunities in AI/ML. Email me directly or use the form.

me@zhachory.com