Own the most critical and complicated part of the project - streams powered services for real-time camera incident detections (person detection, weapon, loitering, tracking and many other types). Production backend is highly loaded, an average rate is about 400M processed requests (messages) per hourResponsibilities:- architecture design, writing docs and specs, tasks creation, guidance for test developers- feature development and delivery to our customers within strict schedule- deployment to cloud environments with no downtime, health monitoring (metrics) and production incidents resolution- work closely with AI team in order to improve detection models and make system smarterAchievements:- improved architecture of backend, made it robust, reliable and highly scalable- rules engine 4X times speedup: real-time frames processing and incident detectionTechnology Stack: Scala, Java, Python, Kafka Streams, Play, Sentry, LaunchDarkly, AWS (EC2, ECS, MSK, CloudWatch, ElastiCache, OpenSearch, CodePipeline, CodeBuild)

Worked in many domains during this time - Insurance, Observability and Data AnalyticsResponsibilities: - designed high-loaded services in Scala with distributed architecture and micro-service approach used- integrated third-party services and APIs into loosely coupled components of our system - deployed applications to the production cloud environments, resolved production incidents - wrote specs, performed architecture design, created tickets- developed state-of-the-art analytics tools to support diverse tasks ranging from ad hoc analysis to production-grade pipelines and workflows for customer applications- partnered with Data Scientists and Analytics Engineers to improve performance and reliability of advanced algorithms Achievements: - led the Trino project, a data-lake enabling searches across data-sources; Trino was delivered and deployed within 5 months; it replaced OpenSearch, saving $1.2 million/year and adding more features; designed ETL pipeline successfully handled about 250M incoming log batches per hour - designed distributed architecture for insurance-analytics from the ground-up; this included the core modules, API, third-party integrations; the cloud-native architecture included high-volume computation capability powered by Akka, Spark, Redis, Kafka and Mongo DB Technology Stack:Scala, Akka, Akka HTTP, Play Framework, Postgres, MongoDB, Kafka, Spark, Specs2, GitLab, Kubernetes, KeyCloak, Cats, Circe, HTTP4s, AWS (S3, Kinesis, MSK, Glue, IAM, CW), Trino

Served on several teams, leading 3 of those projects, to build the infrastructure for online advertising engines and for collecting advertising statsResponsibilities: - designed high-loaded API services in Java/Scala with distributed architecture and micro-service approach used- implemented, and architected, mission-critical and high-loaded components related to pulling stats and managing ad-campaigns in real-time - led end-to-end design of a reporting system powered by an OLAP cube in Scaladeployed incremental upgrades, owned health-monitoring and incidents-resolution for productionAchievements: - led the process of redesigning legacy banner engine to Java with 8x performance improvementthe sophisticated ad-campaign management solution we built was a market-differentiator for NativeAds, that way company signed a few big clients Technology Stack:Java, Scala, Akka, Play Framework, MySQL, Slick, OLAP, AWS S3, Redis, ElasticSearch, ScalaTest

Served on 8+ projects for a diverse client-set Responsibilities: - researched new features, designed and implemented them into software products, this included design, development, bug-fixing and BAU incident resolution- managed a 5-10 developer team, assigning and reviewing their day-to-day tasks- elicited requirements and provided estimates, wrote specs, and designed architecturepeer-reviewed merged and deployed new build to production servers; served as the administrator for high-loaded production servers - interviewed candidate-developers to build a new teamAchievements: - redesigned the architecture of the API, storage and crawler components of PageFreezer project (website and social media archiving tool), added critical features, reduced resources usage- managed 200TB+ data storage with crawled website snapshots and administered an unconventionally large MySQL database (8 GB+) that contained all metadata; performed multiple migrations, with varied scope, without noticeable downtime for our clientsTechnology Stack:Java, J2EE, Spring Framework, Tomcat, Nginx, Rabbit MQ, MySQL, Hibernate, Solr, Gson, JUnit