KULDEEP KUMAR

Led engineering for core fintech and supply-chain platforms, ensuring architecture, SDLC, quality, cost, and reliability across multiple product lines handling millions of daily transactions.

• Architected and delivered unified Accounts Receivable Platform enabling multi-mode payments, automated credit approvals, real-time credit limit enforcement, and reconciliation workflows.
• Introduced structured SDLC practices, design reviews, fault modeling, and observability standards to enhance development processes.
• Established scalable patterns for workflow orchestration, error recovery, and data consistency across financial systems.
• Directed complete rebuild of Udaan's legacy payments stack into a scalable platform, reducing infrastructure costs by 70% through architectural simplification.
• Increased transaction throughput and platform uptime with robust circuit breaking, retry policies, and horizontal scaling frameworks.
• Designed in-house payment router with dynamic routing logic based on real-time gateway success and latency, improving success rate from 90% to 98%.
• Architected high-scale microservices-based trade credit system achieving 99.999% availability and implemented real-time credit scoring pipelines for instant decisioning.

• Designed and developed Autonomous Health Services utilizing Machine Learning for Oracle Autonomous DB.
• Created an integrated interface for real-time health monitoring, diagnosis, and root cause analysis.
• Implemented secure log redaction to enhance data security and compliance.

• Developed deep learning models in TensorFlow for streaming and web traffic classification.
• Implemented VPN traffic classification with Random Forests to enhance detection accuracy.
• Created machine learning-driven device inference behind NAT using TCP fingerprinting.

• Contributed to Port Flexibility feature, enabling virtualization for LANs.
• Developed multiple features in Python and C++ to enhance server virtualization capabilities.