Pooja Desai

Spearheaded the design and scaling of backend systems driving reliability, user safety and real-time insights for millions of active users.

• Architected and launched a first-ever reusable identity platform processing 15M+ lifecycle events across registration, authentication and account workflows, defining shared APIs, state models and event-driven orchestration used across multiple product surfaces while reducing targeted abuse by 60%.

• Leading the architecture of a multi-parent account model, evolving foundational account and relationship schemas to support complex user hierarchies while preserving backward compatibility across existing APIs and production workflows.

• Centralized IP-risk evaluation in a Go service backed by MySQL and Redis, achieving sub-20 ms p95 latency and improving cache hit rate by 35%.

• Consistently met a 99.95% availability SLO across Kubernetes-based services by establishing production

telemetry, actionable alerts, safe rollout controls, incident-response practices and post-incident corrective actions.

• Designed incremental migration and rollout strategies for changes to foundational account workflows, using backward-compatible APIs, feature controls and phased releases to evolve production systems without disrupting existing clients.

• Mentored engineers through design reviews, implementation planning, code reviews and operational decision-making.

Pioneered the architecture and development of distributed microservices and data-driven products for

SMBs, leveraging Temporal workflows, Elasticsearch and cloud-native technologies to deliver scalable and

high-performance backend systems.• Engineered scalable batch-processing pipelines with Temporal workflows, sharding and parallel data

streams, handling large datasets efficiently.

• Drove 8.6% recurring-revenue lift by building gRPC integrations with Google Business Profile APIs and improving SEO workflows.

• Increased listing discovery conversion by 12% and cut query latency by 60% by optimizing Elasticsearch for sub-second search across 10M+ listings.

• Designed reusable OpenAPI-based integration boundaries that exposed core listing-management capabilities across external partners while insulating internal services from partner-specific requirements.

• Mentored engineers on distributed-systems design, concurrency, testing practices and production observability through code reviews and design discussions.

Engineered and led large-scale pricing and promotion systems for JCPenney.com using distributed data technologies, delivering robust and scalable solutions.

• Designed and implemented the organization’s first omnichannel pre-clearance pricing model, ensuring consistent pricing across online and in-store channels.

• Built a scheduled Java batch job that processed high-volume daily price updates in parallel using a thread pool, reducing execution time by 45%, from 8 minutes to 4.4 minutes.

• Modeled complex product and promotion relationships in Neo4j and used Cassandra to store and retrieve high-volume pricing data.

• Built an event-driven pricing pipeline using AWS Kinesis, enabling downstream systems to consume time-sensitive price updates in real time.