Designing Architecture for Real-Time Advertising Platform

Architected and designed a highly scalable, distributed real-time advertising platform for Adikteev. This project involved expanding the system's capacity from handling 150k requests per second to a remarkable 1 million requests per second during peak times.

Shaped the company's long-term technical vision and guided the evolution of its cloud infrastructure on AWS.

Owned the responsibility for continuous infrastructure cost monitoring and optimization.

Collaborated effectively with Product Owners, Data Engineers, Data Scientists, and Infrastructure Engineers to exceed business objectives.

Led a cross-functional team of 6 developers as a Tech-Lead and managed Frontend and Backend Engineering chapters as a Chapter Head.

Established and oversaw interviewing and onboarding practices, enhancing the recruitment process and team integration.

Ensured technological superiority over competitors through innovative practices and solutions.

Technologies Used: JavaScript/NodeJS, Kafka, Scala, Spark Streaming, PrestoDB, Python, Numpy, Scipy, Pandas, Redis, Cassandra, MySQL, RedShift, AWS, Docker, Kubernetes, GitHub, Jenkins, Statsd, Graphite, Grafana, Elasticsearch, Logstash, Kibana.

Challenges and Solutions: The primary challenge was to design and maintain a Demand Side Platform software capable of scaling significantly while maintaining high availability and network resilience. Through strategic planning and implementation of cutting-edge technologies like Scala, Spark Streaming, and Kubernetes, the platform's capacity was successfully expanded to handle up to 1 million requests per second.

Impact: This enhancement in system capabilities not only improved platform stability but also provided greater visibility into its state, simplifying issue resolution and significantly boosting operational efficiency.

Lessons Learned: Gained extensive knowledge in scalability, efficient software writing and optimization, availability, and network resilience. This experience has been instrumental in understanding the complexities of high-load systems and the importance of continuous innovation in technology.