Abhay
Raghuwanshi

System Design

• Architected an asynchronous, event-driven system for daily batch operations, integrating file processing, APIs, and webhooks to improve throughput and reliability
• Designed and deployed a workflow orchestration engine in Azure using Temporal, achieving 900 workflows per minute
• Designed a multi-invoice reporting system using Strategy and Abstract Factory patterns, abstracting invoice types (InvoiceA, InvoiceB, etc.) behind a common interface to support extensible report generation without modifying core logic

Software Development

• Developed and deployed a RESTful API for VAT computation and invoicing across financial transactions, accounting for multiple invoice types and billing frequencies; implemented a pipeline-style processing model where each billing step is an interface-driven handler, allowing independent sub-steps to be composed, reordered, or extended without touching core scheduling logic
• Implemented post-sweep statement generation service using dual RabbitMQ connections
• Replaced leader-based Spring scheduler with distributed Quartz Scheduler; built a custom annotation to handle job persistence and feature-flag logging, enabling zero-friction migration across 8 microservices
• Applied retry, circuit breaker, and rate limiting patterns across APIs using Resilience4j, improving fault tolerance and preventing cascading failures in downstream services

Data Migration & Legacy Modernization

• Automated migration of 3,000+ cash pool customers from Oracle to PostgreSQL via Jenkins, reducing manual onboarding by 80%
• Reduced data migration time by 2 hours per batch with optimized automation scripts
• Migrated legacy CTI authentication to enterprise OAuth 2.0 on WebLogic with zero data loss

DevOps, Cloud & Automation

• Created and managed CI/CD pipelines using GitHub Actions for 10+ microservices
• Integrated HashiCorp Vault for centralized secrets management across all services
• Implemented graceful shutdown for Spring Boot services in Kubernetes using Actuator endpoints
• Set up observability stack with Filebeat, APM, and Metricbeat on Azure; configured deployment-aware APM settings and condition-based email alerts for proactive incident detection
• Built Airflow DAGs with automated DB queries to run morning production health checks across multiple countries' EOD processes, replacing manual support routines with scheduled validation and alerting

• Led research on the US housing market analyzing 100+ demand and supply factors impacting house prices
• Used Python (Pandas, Tableau) to clean and visualize data — reports increased user traffic by 10%
• Built ML and deep learning models to predict house prices for investment insights

• Developed order management application with role-based dashboard
• Optimized performance with pagination and filters, reducing search time