Abhishek Namdev Sawant

About

Backend/platform engineer with payments and insurance domain experience (Cognizant), owning high-volume production systems end-to-end. Shipped a Spring Boot payment tokenization service with strong observability and phased migration that cut transaction latency ~30% and improved payment efficiency ~25%. Also productionized an ML-driven financial health/risk analytics pipeline with near real-time dashboards across 70+ schools, emphasizing interpretability, data quality, and drift monitoring.

Key Strengths

• Owned end-to-end production delivery (frontend/backend/data/infra) for payment tokenization in high-volume microservices
• Measured performance improvements: ~30% lower transaction latency and ~25% higher payment efficiency
• Strong production readiness practices: unit/integration/load testing plus logging/metrics/alerts and runbooks
• Incremental modernization of legacy payments to microservices using API gateways, contracts, feature flags, phased traffic, and fallbacks
• Built and productionized ML risk/health classification system with ETL + dashboards across 70+ schools
• Pragmatic ML approach: interpretability, confidence scores, cross-validation, drift monitoring, and data quality validation
• Effective stakeholder partnership to translate ambiguous goals into roadmaps and phased delivery (insurance processing improvements)
• Able to operate across multiple services/teams with minimal oversight while shipping incrementally under tight timelines
• Built production-grade payment processing backend on Java/Spring Boot microservices on GCP
• Debugged and resolved peak-hour latency/timeouts via SQL optimization, indexing, and reducing synchronous coupling (30% latency reduction)
• Designed resilient event-driven, stateful multi-step transaction workflow with idempotency, retries (exponential backoff), and manual escalation paths
• Relational data modeling for payments with audit-friendly event/state tables and integrity constraints
• Applied ML delivery end-to-end (data pipelines to Power BI dashboards) for financial health/capacity prediction across 74 schools (~69–71% accuracy)
• Strong data quality guardrails: preprocessing pipelines, schema/input validation, and safe fallbacks to prevent bad predictions

Reference Highlights

• High overall effectiveness (9/10)
• Strong performance optimization impact
• Improved microservices communication latency via redesign
• Designed reliable multi-step payment workflows across distributed services
• Strong debugging under peak-load conditions
• Log analysis to identify root causes
• Optimized critical SQL queries to improve response times
• Pragmatic use of async calls for non-critical steps
• Clean, modular, readable Python code
• Strong refactoring skills while preserving correctness
• Focuses on root causes rather than patch fixes
• Thinks through failure modes and edge cases in payment systems
• Boosted team productivity and code quality; reduced boilerplate
• Adapts well to frequent requirement changes
• Takes ownership of own and team tasks end-to-end
• Open to client feedback
• Delivered smoothly under tight deadlines and estimate constraints

Experience

Education

Skills

Languages