Kyle S. Smith

Led architecture and delivery of a cloud-native clinical operations platform using Python/FastAPI, PostgreSQL, and React/TypeScript. Improved end-to-end workflow throughput by ~35% through streamlined UI and API design. Introduced a microservices approach on Docker and Kubernetes (EKS) with standardized service templates to reduce new-service setup time from days to hours. Implemented secure OAuth2/OIDC/JWT authentication and fine-grained RBAC aligned to least privilege and audit requirements. Built event-driven integrations for asynchronous processing using Kafka-style pub/sub patterns and idempotent consumers, reducing downstream processing latency by ~40% during peak periods. Implemented an AI-assisted triage and summarization workflow with LLM APIs and human-in-the-loop review, reducing manual documentation time by ~25% while maintaining guardrails and traceability. Rolled out observability with OpenTelemetry and Prometheus/Grafana, improving MTTR by ~30% via actionable dashboards

Owned end-to-end delivery of internal platform features across Next.js and TypeScript APIs on AWS Lambda/API Gateway and infrastructure, reducing manual operational handoffs by ~40% through workflow automation. Built an integration layer using REST, OAuth 2.0 (PKCE/client credentials) and webhooks, normalizing inconsistent third-party payloads into a stable domain model and cutting partner onboarding time from weeks to days. Designed event-driven automation with SQS and DynamoDB-backed state tracking, using idempotent handlers to orchestrate multi-step business processes with retries, deduplication, and auditability. Delivered internal developer tooling: a Node.js/TypeScript CLI for token management, sandbox testing, and replaying webhook events to improve throughput and reduce production integration defects. Implemented pragmatic AI automation using LLM APIs (Anthropic/OpenAI) for document/content extraction and routing, achieving ~60% reduction in manual triage while adding human-in-

Led modernization of Azure-hosted internal portals and APIs using C#/NET (ASP.NET Core Web API), improving page/API response times by ~35% through query tuning, caching, and payload shaping. Built and maintained integration services for clinical/operational workflows using REST, Azure Functions, and Azure API Management, reducing manual handoffs and speeding downstream processing by ~25%. Implemented role-based access control and authentication flows using OAuth2/OIDC, Azure Key Vault, and least-privilege service principals; added audit logging to support compliance for sensitive data. Developed UI enhancements for staff-facing tools with React, TypeScript, HTML/CSS, and accessibility checks, reducing usability-related tickets by ~20%. Created automation for intake, triage, and routing of operational requests using Python scripts and Azure Functions, cutting average time-to-triage from hours to minutes. Introduced LLM-assisted automation for summarizing tickets and extracting structure