Tanvik Reddy Kotha

About

Backend engineer who has built high-throughput analytics and fraud-detection systems, combining Python/Flask + Celery/RabbitMQ with strong PostgreSQL performance tuning (indexing, partitioning, EXPLAIN ANALYZE). Has production experience integrating ML inference (scikit-learn/TensorFlow → TensorFlow Lite) into Spring Boot microservices with caching and model versioning, plus designing secure multi-tenant architectures using JWT-based tenant routing and PostgreSQL RBAC/RLS.

Experience

Education

Key Strengths

• Designed and deployed high-throughput Flask analytics API with async processing (Celery + RabbitMQ) on AWS ECS
• Diagnosed and resolved RabbitMQ queue buildup via worker concurrency tuning and splitting heavy jobs into parallel subtasks
• Improved performance and memory usage with pagination and pushing aggregation to SQL
• PostgreSQL/SQLAlchemy scaling: indexing, partitioning by date, lazy loading, and using EXPLAIN ANALYZE to find bottlenecks
• Integrated ML fraud detection into production services with TensorFlow Lite deployment, caching, and model versioning/rollback
• Multi-tenant data isolation strategy using tenant-specific schemas vs dedicated databases, JWT tenant routing, and RBAC/RLS enforcement
• Reduced scoring latency in a millions-of-transactions/day environment using task priorities and Redis TTL caching