Jeffrey Zeng

Led backend architecture for a distributed clinical AI platform processing millions of clinical events and AI workflow requests. Architected ambient clinical documentation and nurse handoff summarization workflows using Gemini/MedLM and FHIR interoperability.

Architected Java/Kotlin platform services and declarative service specifications that standardized compute, databases, queues, caches, IAM, autoscaling, monitoring, and deployment policies across AWS and Kubernetes environments. Built self-service infrastructure workflows that reduced new-service setup time from hours to under an hour.

Built Java and Python backend services powering marketplace matching across category, geography, availability, and eligibility constraints for high-volume consumer service requests. Redesigned candidate retrieval and ranking pipelines using indexed geo/category lookups, Redis caching, parallel enrichment, and asynchronous scoring, reducing p95 matching latency by approximately 40% under peak traff

Developed C#/.NET backend services for healthcare data ingestion, normalization, provider integrations, scheduling workflows, and patient-facing applications on Azure. Designed asynchronous integration workflows with Azure Service Bus using idempotent processing, retry policies, poison-message isolation, and replay support.

Bachelor of Arts in Applied Mathematics from the University of California, Berkeley, completed in 2015.