Sukesh Sadhu

About

Backend/data engineer focused on fintech payments and fraud systems, owning real-time Kafka-based reconciliation pipelines end-to-end (~13k tx/day). Built audit-friendly validation/reconciliation (SQL + Python), kept lag to seconds, and cut errors ~20%, while also shipping Spring Boot APIs with Redis caching and strong idempotency/versioning. Has early-stage startup experience standing up payment services on AWS with Docker + GitHub Actions and production monitoring/incident handling.

Key Strengths

• Owned end-to-end real-time payment reconciliation pipeline (Kafka → validation → Python/SQL transforms → APIs)
• Handled ~13,000 payment transactions/day with processing lag kept under a couple seconds
• Improved reliability by enforcing ordering via Kafka partitioning by transaction ID and idempotent consumers
• Reduced pipeline errors by ~20% through schema validation, idempotency checks, and reconciliation logic
• Built SQL-based reconciliation checks across authorization/capture/settlement stages to detect missing/inconsistent records
• Implemented DLQs, retries/backoff, and time-window buffering to reduce noise and prevent pipeline blockage
• Shipped Spring Boot APIs with Redis caching and query/index optimizations to address slow queries on large tables
• Stood up early-stage payment authorization service with Docker, GitHub Actions CI/CD, AWS deployment, and monitoring from day one

Experience

Education

Skills

Languages