Oct 2023 – Present · 2 yrs 8 mos · Gurugram, Haryana, India · On-site

• Built ETL pipelines using Spark and Hadoop for the DNS Analytics Platform, processing 100M+ daily DNS/NAT/usage logs to enable real-time user classification.
• Automated workflows using Apache Airflow to reduce manual intervention by 80% and improve observability across DNS analytics pipelines.
• Created Grafana dashboards for the DNS platform to monitor traffic trends, detect anomalies, and track system health in real time.
• Developed a Change Data Capture (CDC) microservice using Spring Boot and Kafka to sync 300M+ records from Aerospike to Oracle with zero downtime.
• Maintained live data consistency during large-scale migrations by enabling real-time sync between source and target systems.
• Converted and deployed LightGBM and TwoTower models to ONNX, ensembling pre/post-processing within Triton Inference Server.
• Served recommendation model predictions via low-latency (<50ms) gRPC/HTTP endpoints on OpenShift.
• Built a Jenkins CI/CD pipeline with dual Git and JIRA triggers to automate deployments across SIT, UAT, and Prod environments.
• Migrated the internal container registry from Nexus to Quay, improving deployment speed and reliability with zero downtime.
• Integrated Darts 2.0 into 30+ ETL pipelines to deliver fault-tolerant, high-availability reporting across CRM systems.
• Engineered key modules for the Unified Customer Management (UCM) Platform, Airtel’s centralized data hub serving 300M+ users and enabling cross-domain CRM unification.