Sahithi Mogudala

About

Full-stack engineer with enterprise experience at Metasystems Inc. (and Qualcomm) building high-traffic, security-sensitive systems—owned a secure transaction processing module end-to-end using Java/Spring Boot, Python/Django, and React. Strong AWS production operations (EKS/ECS/Lambda/RDS/DynamoDB) with IaC (Terraform/CloudFormation), observability, and reliability patterns; also delivered resilient ETL/integration pipelines with idempotency/retries/backfills and achieved a 50% deployment-time reduction through CI/CD and modular refactoring.

Experience

Education

Key Strengths

• Owned secure transaction processing feature end-to-end (frontend, backend, deployment/ops) for high-traffic enterprise payment platform
• Designed and built scalable microservices with secure auth (Spring Security + JWT RBAC) and audit logging
• Production AWS operations experience (EKS/ECS, autoscaling, load balancing, WAF/Shield) with observability via CloudWatch and structured logging
• Performance optimization across stack (Postgres query tuning/indexing, caching, pagination/batching) with measured latency/API efficiency improvements
• High-ambiguity delivery: designed flexible APIs/modular data models and shipped on schedule with evolving requirements
• Maintainability/scalability improvements via refactoring into modular components, enforcing standards/TDD; reduced deployment time by 50%
• Built resilient data pipelines/integrations with idempotency, retries, backfills, reconciliation, and alerting
• Designed stateless JWT auth with secure password hashing (bcrypt)
• SQLAlchemy/PostgreSQL performance tuning using EXPLAIN ANALYZE, indexing, partitioning, and connection pooling
• Resolved concurrency issues (query locking) by tuning transaction isolation and pooling
• Integrated TensorFlow ML inference into Flask APIs with Redis caching and batching to reduce latency
• Containerized and scaled services with Docker + Kubernetes; isolated model/runtime dependencies via separate images/pods
• Implemented blue-green deployments for zero-downtime model updates
• Built multi-tenant data isolation using tenant ID middleware/JWT claims and ORM-level tenant filters
• Improved API responsiveness by offloading work to Celery + Redis; reported ~30% response-time improvement