Software Engineer, Dun & Bradstreet, San Jose, CA Company profiles, B2B. Data Engineering, Fuzzy search, Clustering

• Increased search hit rate by 4% by developing a Spelling correction module. Applied Trie as

BloomFilters, Count-min-sketch for language models. 7M unique words, 55M pairs, 55M triples.

• Researched and developed a de-duplication method for 600M company profiles. Locality sensitive hashing, MinHash, Trie as BloomFilters for Hamming distance. Found 7M exact and fuzzy duplicates.

• Implemented Verification and Identification services that match web domains to companies. 60M

websites scale.

Technologies: Scala, Python, Postgres, ML, AWS

Software Engineer, Piano Media, Izhevsk, Russia

A platform for paywall and subscription business. Distributed systems, ML, Reports, DSL, Streaming.

200 Tb data in total, 5000rpsavg, 8000rpspeak daily.

• Launched Likely-to-Subscribe and Likely-to-Churn ML projects, 10M users

• Led and built a project that provides various types of analytic reports, real-time, on-demand,

customizable reports based on DSL. Connected Clojure-based DSL with Spark and BigQuery.

• Implemented User Segmentation Engine, 100M users, used Bloom filter as KV-store Technologies: Scala, Clojure, Unix, Spark, HDFS, DSL, AWS, Akka, Kafka, Catboost, BigQuery

Researcher, Senior Algorithm Engineer, Huawei Research, Moscow, Russia

Nested data parallelism model, distributed systems, map-reduce, optimization compiler, meta-

programming and code generation, domain-specific languages

• Implemented Spark backend for nested data-parallelism model

• Researched and developed efficient custom Spark RDD’s for NDP-primitives Technologies: Scala, Spark, Distributed Programming, C++