Designed and implemented event driven microservices exposed through REST APIs in a Service Oriented Architecture (SOA), sustaining a steady 100 to 1K RPS peak and publishing 1K+ lifecycle events per day to keep downstream systems consistent, deployed on Kubernetes in GCP for autoscaling and resilience.

Delivered an integration layer for smart sensor emulators using pub/sub triggers and validation rules, automated 50+ real time environmental tests and cut manual QA effort 80%, saving 30+ hours per week and improving internal product reliability.

Developed a full stack door configurator adopted by 500+ distributors across 100+ partner accounts, cutting manual quote time by 90% percent per order.

Modeled and tuned database schemas with partitioning, indexing and pre-aggregation and optimized queries, bringing analytics latency down from 3 minutes to 12 seconds and enabling business teams to self-serve portfolio and price variance reports that previously required engineering support.

Owned production readiness for new services by standardizing monitors, alerts and incident runbooks, led code reviews and sprint rituals for 2 teams and kept availability and sprint commitments stable during rollouts.

Deployed 100+ microservices for real time ingestion and external APIs with batching, autoscaling and circuit breakers, improved throughput by 40% and stabilized availability for high volume commerce data.

Partnered with a data science team to convert raw streams into production features for demand forecasting and price optimization, which upgraded planning from descriptive to predictive.

Ran the rotating on-call, maintained 99.95% monthly availability, reduced median MTTR from 45 to 20 minutes with runbooks and alert tuning and drove blameless postmortems that removed repeat incidents by 50%.

Built an event driven Customer Data Platform ingesting 500K+ daily behavioral and offline events using identity stitching, enabled richer segmentation, and delivered a 40% engagement jump and 25% sales uplift.

Led a 7 person delivery team and rollout, trained internal and client users and set governance and SLAs so the platform scaled without constant engineering support.

Governed REST APIs on profile data with RBAC, rate limits and audit logs, which let marketing activate high intent segments in near real time while meeting client data policies.

Built an ad copy generation platform that scaled to 200K+ copies per day with asynchronous queues and template ranking, raised creative throughput by 60% and drove up client revenue by 25%.

Removed bottlenecks in legacy ad tech services using caching and parallelization, improved throughput by 90% and shortened campaign delivery timelines with higher customer satisfaction.

Unified internal optimization engines and external channels via REST APIs in a Service Oriented Architecture (SOA), which accelerated onboarding of new advertisers.

This project was an attempt to develop smart layer in between OS and kernel which aims to improve the storage utilization by removing redundant data from the real time storage requests.

Implemented on Device-Mapper framework to get control of real time I/O requests.

Used SHA-256 hashing algorithm to uniquely identify different and similar blocks being read.

The complete system is being designed and developed as a part of the University BE Project.