MEGHANSH NAGULA

About

Full-stack engineer with recent JPMorgan experience building GPT-4-powered customer sentiment/feedback analytics products (Next.js 14 App Router + FastAPI + Postgres) and owning them post-launch with CloudWatch/Datadog observability. Also implemented Temporal-based transaction reconciliation workflows with strong reliability patterns (idempotency, retries, DLQ, versioning) and has prior high-scale healthcare dashboard experience at Optum.

Experience

Education

Key Strengths

• Shipped end-to-end Next.js 14 + TypeScript customer insights dashboard (UI through deployment) with post-launch monitoring/optimization ownership
• Balanced freshness vs performance using Next.js caching strategies (force-cache/no-store/revalidate) plus React Query for delta updates
• Improved perceived load time ~40% using React 18 streaming/Suspense and server-heavy rendering strategy
• Reduced unnecessary re-renders ~70% via targeted memoization (React.memo/useMemo/useCallback) validated with profiling
• Designed Postgres analytics schema with constraints and indexes (composite + GIN) for feedback/sentiment tagging use case
• Cut analytics query latency from ~3s to <100ms using pre-aggregated views and scheduled refresh; validated with EXPLAIN ANALYZE and pgbench
• Built durable Temporal workflow for transaction reconciliation with idempotent activities, retry/backoff, DLQ pattern, and workflow versioning
• Pragmatic architecture tradeoffs: shipped GPT integration embedded for speed, later extracted to microservice with Redis queue when scale demanded