• Designed a serverless replacement for legacy Solr-based search using Node.js, AWS Lambda, and Amazon RDS

(MySQL), eliminating production failures and reducing monthly infrastructure cost by ~55%

• Developed data ingestion pipelines using Apache Airflow, AWS Glue, and PySpark with Snowflake as the source

system, contributing to a 30% improvement in indexing freshness for search datasets

• Tuned Snowflake SQL queries and PySpark transformations, decreasing AWS Glue job runtime from ~30 minutes to

~4 minutes per table, achieving a 7.5× execution improvement

• Migrated CI/CD workflows to AWS CodePipeline using AWS CDK, improved reliability & reduced release cycle by 20%

• Enhanced AWS Lambda observability using CloudWatch logs, metrics, and alarms, lowering monitoring cost

contribution by ~15% while maintaining alert coverage across critical services

• Engineered a real-time vehicle and tyre tracking platform for European agricultural fleets using Svelte and Node.js

(TypeScript) with PostgreSQL, Redis, and RabbitMQ, increasing tracking accuracy coverage

• Developed a new backend service to support integration of a new Teltonika GPS device type, enabling onboarding of

additional customers and achieving a ~7% increase in device adoption

• Analyzed Redis caching strategies in API layers, and decreased average response time by 25% for tracking endpoints

• Built a real-time map-based tracking system using WebSockets and APIs, monitoring device service needs and

engineer locations, improving assignment decisions, and boosting service response rates by ~50%

• Refactored legacy React codebase from ES5 to ES6 standards, leading to a 30% reduction in code complexity and

improving cross-browser compatibility across major supported environments