Pratham Rathour

Led end-to-end migration of RefAdmin from legacy on-prem monolithic system to cloud-native microservices architecture (application + database), eliminating dependency on legacy infrastructure and saving $25,000 annually; consolidated user management into a unified platform integrating real-time data streams ensuring strong consistency. Re-architected data synchronization pipeline eliminating 20-minute replication lag, enabling real-time propagation of user roles and permissions and reducing role-update incidents from ∼2/month to 0 defects over 6 months, significantly improving system reliability. Owned and optimized client onboarding and offboarding workflows achieving 100% accuracy and eliminating production issues; simultaneously resolved multiple Sev1/Sev2 incidents across distributed services and executed critical production scripts with 100% success rate.

Member of Technical Staff 2 at Nielsen (2024-05 – Present)

• Led end-to-end migration of RefAdmin from legacy on-prem monolithic system to cloud-native microservices architecture (application + database), eliminating dependency on legacy infrastructure and saving $25,000 annually; consolidated user management into a unified platform integrating real-time data streams ensuring strong consistency.
• Re-architected data synchronization pipeline eliminating 20-minute replication lag, enabling real-time propagation of user roles and permissions and reducing role-update incidents from ∼2/month to 0 defects over 6 months, significantly improving system reliability.
• Owned and optimized client onboarding and offboarding workflows achieving 100% accuracy and eliminating production issues; simultaneously resolved multiple Sev1/Sev2 incidents across distributed services and executed critical production scripts with 100% success rate.
• Optimized dictionary export pipeline reducing latency by 70–80% for large datasets and improving storage utilization by 15–20%, leading to enhanced performance and system stability under high-load scenarios.
• Engineered event-driven asset tracking system using Amazon SQS enabling real-time monitoring and scalable event processing; contributed to technical hiring, mentored new graduates on system design and architecture, and led L1/L2 code reviews ensuring high-quality engineering standards.

Member of Technical Staff & Intern at Byju's (2023-01 – 2024-05)

• Played a key role in building OneDesk, a scalable microservices-based platform replacing Zendesk, enabling 10,000+ active users and driving annual cost savings exceeding 10+ crores, significantly reducing operational expenses.
• Designed and implemented Bulk Case Upload system processing 5000+ cases/min through optimized CSV ingestion and backend architecture, significantly improving operational efficiency and reducing manual overhead.
• Led large-scale migration of 2M+ cases (including detailed history and metadata) from Salesforce to OneDesk with throughput of 10,000 cases/min, ensuring high reliability, zero data loss, and seamless transition of critical business data.
• Improved API performance by reducing latency up to 65% through optimization of business logic, caching strategies, and database interactions, significantly enhancing user experience.
• Optimized MongoDB queries using indexing and query tuning techniques improving data retrieval efficiency and reducing response times under high-load conditions; also migrated Zendesk triggers and workflows via custom scripts ensuring smooth system transition.
• Built comprehensive observability and monitoring using Prometheus and Grafana by defining custom metrics for latency, error rates, and service health, significantly improving system reliability, debugging efficiency, and production visibility.

B.E. in Electronics and Computer Engineering – Thapar Institute of Engineering and Technology (2019-2023)