Dmytro Pazykov

About Dmytro Pazykov

Software Development Engineer with 5+ years at Amazon, owning end-to-end design, development, and operations of scalable, AWS-hosted applications powering device management for Ring and Kindle ecosystems. Engineered high-throughput solutions processing millions of daily device events, leveraging Lambda for serverless orchestration, SQS for reliable queuing, DynamoDB for low-latency NoSQL storage, S3 for durable data handling, and CloudWatch for proactive alerting and observability—delivering 99.99% uptime and 50-70% reductions in operational latency/costs. Deep expertise in distributed systems, microservices, and event-driven architectures using Java/Python. Battle-tested on-call leader with cross-team impact. Passionate about clean, resilient code at planetary scale. Eager to architect innovative services at forward-thinking tech leaders.LanguagesJava 17 (primary), Python 3, TypeScript/JavaScript.Backend & Distributed SystemsMicroservices, event-driven architecture, REST/gRPC API design, idempotency, eventual consistency, CQRS patternsAWS (Production-scale)Lambda, DynamoDB (GSI/LSI, streams, transactions), S3, SQS/SNS, Step Functions, API Gateway, ECS/Fargate, CloudWatch (metrics, alarms, X-Ray tracing), IAM, Kinesis (occasionally)Databases & StorageDynamoDB, Aurora PostgreSQL, Redis/ElastiCache, S3 lifecycle policiesInfrastructure & DevOpsDocker, Terraform (basic), Jenkins, CodePipeline, Spinnaker, Git/GitHub, Bitbucket Pipelines, Gradle, observability & alertingKey Tools & PracticesHigh-throughput/low-latency systems, on-call rotation, chaos engineering principles, metrics-driven development, cost optimization

Experience

1. Software Engineer

   Amazon

   May 2024 — Present · Seattle, WA, US

   Owned and led a greenfield internal service from 0 → production, building a globally available, multi-region backend that securely reads physical Kindle devices at scale and delivers validation results to downstream teams- Designed and implemented high-throughput, event-driven architecture using AWS Lambda, Step Functions, DynamoDB (with streams and global tables), SQS/SNS, and S3, achieving 99.99% availability across US, EU, and FE regions- Built resilient multi-region data replication and failover logic, enabling seamless region switchover with zero customer impact during outages- Developed core backend services in Java 17 and Python, exposing clean REST/gRPC APIs consumed by 20+ internal services and processing millions of device events daily- Engineered dynamic device routing, scheduling, and resource allocation system that improved peak throughput by 4× and reduced average validation latency by 70%- Integrated end-to-end observability with CloudWatch, X-Ray, custom metrics, and automated alerting, cutting MTTR from hours to minutes and eliminating entire classes of production incidents- Embedded generative AI-driven predictive diagnostics into the IoT pipeline, proactively identifying device anomalies before they impacted validation success rate- Owned CI/CD pipelines (Jenkins + CodePipeline), enforcing trunk-based development and achieving multiple production deployments per day with zero downtime.

Education

• Petro Mohyla Black Sea State University

  Bachelor's degree, Computer Science

  2011 — 2015