Megha Jain

Owned the API layer for the Benefits & Enrollment domain — redesigned request handling, introduced response caching, and tuned connection pooling, cutting p99 latency by 40% on endpoints serving enterprise-scale traffic. Drove a full SQL performance overhaul: redesigned index strategies, rewrote long-running queries, and normalized schemas on the highest-traffic benefit plan tables, yielding ~50% faster reads. Built automated reporting pipelines that replaced a suite of recurring manual workflows, reducing ops team effort by 90% and enabling same-session business analytics previously unavailable.

Established a Datadog observability layer from the ground up for the Benefits domain — APM traces, custom alert policies, and structured logging that reduced mean-time-to-detect production issues by ~60%. Led P1/P2 incident response across enrollment and pricing workflows; authored post-mortems and drove permanent fixes that brought regressions in these areas to zero over two release cycles. Architected the migration of legacy benefit plan creation from a monolith to independent RESTful services backed by a React frontend — decoupling release cycles and improving admin configurability.

Introduced systematic input validation and error-boundary patterns on EOI document and pricing flows, eliminating a recurring class of runtime failures that previously required manual intervention. Participated in technical design reviews and cross-team planning sessions with QA, Product, and Dev Ops; reviewed code across the team with a focus on correctness and long-term maintainability.

Designed the microservices architecture for a new product line — defined service boundaries, API contracts, and data ownership; services were independently deployable from day one. Built async event-driven reporting pipelines using Rabbit MQ, decoupling data producers from consumers and reducing stakeholder reporting latency from hours to near-real-time.

Rewrote Mongo DB query layer using compound indexes and optimized aggregation pipelines, reducing p95 query time by ~50% across the most frequently accessed collections. Served as sole backend owner of an AI-powered document generation platform — designed the generation API for concurrency, added rate-limiting, and built an analytics module that tracked conversion and engagement metrics used in product decisions.

Built a scheduled domain monitoring service with multi-channel alerting; maintained zero domain-related outages across 12 months of production operation. Replaced manual project status workflows with a Telegram Bot integration, cutting recurring reporting effort and giving stakeholders real-time visibility into delivery status.

Built modular Spring Boot services for capital markets applications, improving deployment independence across product verticals. Contributed across the full delivery lifecycle — requirements, development, testing, and production support — gaining practical exposure to enterprise-grade software operations.