Bangalore

• Led design and implementation of a Relation-Based Access Control (ReBAC) model for agentic, multi-vendor data protection and cybersecurity systems (patent filed).
• Built an AI evaluation framework (LLM-as-a-Judge) with automated scoring pipelines and benchmark datasets to measure model performance at scale.
• Developed an integration platform for ServiceNow and a LangGraph-based troubleshooting solution.
• Designed canonical complex data models spanning multiple data protection vendors to ensure consistent schema and interoperability.
• Architected cloud infrastructure and CI/CD pipelines to automate infrastructure deployments and code evaluation workflows.
• Led the Azure DevOps to Jira migration; designed the full Jira system architecture, including workflows, taxonomy, Advanced Roadmaps and the cross-team coordination model.
• Managed 8 engineers across frontend, backend, AI and DevOps; provided architecture direction and technical mentoring.
• Served as Scrum Master for a ~25-person team; facilitated sprint ceremonies and proactively removed delivery blockers.

Singapore

• Built ML data pipelines integrating with building management systems, enabling automated decision-making and cutting processing time by 50%.
• Implemented Brick Schema to enhance semantic metadata storage for building management systems, improving interoperability and maintainability.
• Collaborated with cross-functional teams to streamline data-driven decision-making, improving predictive analytics models and fault detection accuracy by 25%.
• Migrated microservices to Kubernetes and streamlined CI/CD with GitLab, achieving a 99.99% SLA and 40% faster deployments.
• Developed a high-throughput, bi-directional data pipeline in Golang and Python, achieving a 100x increase in performance with zero data loss.
• Enhanced database reliability by moving from a standalone InfluxDB to a highly available, replicated setup, ensuring 99.99% uptime.
• Improved IoT device observability by integrating Prometheus and Grafana, reducing downtime by 20% and increasing data accuracy by 25%.

Bengaluru

• Developed and optimised a highly scalable IoT data pipeline using Python, RabbitMQ and MongoDB, enabling real-time predictive maintenance and a 20x increase in scalability.
• Implemented Change Data Capture (CDC) for MongoDB between edge and cloud databases to sync IoT data from edge devices to cloud infrastructure.
• Built advanced analytics and data management modules for a deep learning system processing 300 parts per minute, significantly improving defect detection efficiency and accuracy.
• Designed an edge-to-cloud database sync using Python and AWS S3 with automated backups, high availability and seamless disaster recovery — reducing data recovery time by 50%.