Sep 2023 – Present · 2 yrs 8 mos · Pune, Maharashtra, India

• Leading architecture and delivery of MPL’s real-time data platform processing 2.6M+ events/sec during peak concurrency, powering fraud detection, personalization, and rewards for 100M+ users.
• Designed and owned end-to-end a unified event ingestion pipeline (Kafka → BigQuery) handling 700+ unique event types, reducing ingestion latency from 30 minutes to sub-second at 2.6M EPS peak.
• > Architected a high-performance real-time Feature Store serving 1,000+ user features with 100K RPM throughput and p95 latency <50ms.
• > Built a low-latency streaming ML inference platform achieving ~500ms average inference at 2,000 EPS scale.
• > Delivered a real-time IFTTT-style rewards engine supporting 900+ dynamic use-cases with <100ms latency at 1,200 EPS.
• > Designed and scaled real-time user segmentation microservices and A/B experimentation platform processing 1,500 EPS, enabling rapid product iteration.
• > Established platform-wide observability, schema governance, and reliability practices; mentored 10+ engineers with multiple promotions under my guidance.
• Tech stack: Java, Kafka, Flink, GRPC, Redis, Cassandra, Postgres, MongoDB, BigQuery, GCP, Kubernetes.

Feb 2022 – Sep 2023 · 1 yr 7 mos · Pune, Maharashtra, India

• Played a key role in building and scaling MPL’s next-generation data platform powering real-money gaming for millions of users across India.
• Designed and led development of critical features on a unified data platform supporting both batch and real-time (Kafka+ Java )pipelines, processing terabytes of gaming, transaction, and user-behavior data daily.
• Architected low-latency, fault-tolerant data ingestion and transformation systems that reduced end-to-end data latency by 60% .
• Championed observability and reliability initiatives: introduced schema registry governance, real-time monitoring dashboards, and automated data quality frameworks, cutting production incidents related to data pipelines by 75%.
• Collaborated closely with Data Science, Product, and Risk teams to deliver high-impact features including real-time fraud detection streams and personalized recommendation data feeds.
• Mentored junior engineers and drove best practices in code quality, testing, and CI/CD, raising overall team velocity and engineering standards.
• Result: Enabled faster, more reliable data-driven decisions across the organization while supporting MPL’s hyper-growth phase.