AadarshKumar Vanga

About

Backend/ML systems engineer who built a Flask + PostgreSQL internal ticketing platform and demonstrates strong database/ORM performance depth (indexes, partitioning, RLS multi-tenancy). Notably optimized a high-throughput attachment OCR/embedding pipeline with batching, deduplication, and Redis caching, cutting median latency from 45s to 10s and reducing worker cost by 35% while increasing throughput 4x.

Experience

Education

Key Strengths

• Built and maintained Flask-based internal ticketing backend with JWT auth, validation, and clean ORM patterns
• Deep SQLAlchemy relationship modeling (cascade behavior, circular import mitigation, base model patterns)
• Performance tuning for ORM/database (addressed N+1 queries; selective loading strategies)
• PostgreSQL scalability expertise (index design, partial indexes, partitioning, DB-level constraints)
• Multi-tenant data isolation design using tenant scoping plus Postgres Row-Level Security
• High-throughput pipeline optimization with measurable impact (median latency 45s→10s; p95 <30s; 4x throughput; 35% worker cost reduction)
• Production-minded background processing design (batching, idempotency, deduplication, Redis caching, CPU/GPU pool separation)
• Broad ML systems integration experience across online, batch, and streaming inference architectures