Prasad Shegokar

Built a high-performance Collector And Tagger in Rust, using the Outbox Pattern to process over 2M+ packets/sec in a distributed system. Integrated Kafka Streams, Schema Registry, and Connect to enable versioned, scalable data pipelines across services. Led design and delivery of the ONT Speed Test platform in Java, covering HLD/LLD and cross-functional execution. Resolved architectural issues in CNAP provisioning, improving backend reliability and throughput. Developed deep expertise in AXOS architecture, RPC-based orchestration, and device provisioning. Mentored junior engineers, delegated tasks, and conducted code reviews to ensure high-quality outputs. Documented workflows and led onboarding sessions to accelerate team ramp-up.

Architected and optimized a high-throughput notification backend in Java and Python, capable of handling 200K+ push notifications/sec, improving user engagement and system responsiveness. Streamlined deployments using Terraform and Docker, enabling fast, reliable delivery of Java/Python backend services. Developed real-time, event-driven pipelines with Redis, Kafka, and Kinesis for scalable async processing. Refactored middleware in Java to support auto-scaling, multi-client workflows, and high concurrency.

Migrated Mainframe applications to Java, ensuring cloud compatibility and scalability. Refactored and containerized Spring Boot apps, deploying on AWS for enhanced scalability and performance. Optimized low-level design, improving performance by segregating components.

Built a generic PDF parser using Advanced Java for processing account statements, unlocking new business opportunities. Developed a micro-application to reconcile multi-bank data with HSBC systems. Optimized queries and resolved performance issues, boosting efficiency 20x.