San Diego, California, United States
• Built production LLM report-generation workflows for commercial, government-facing, and internal users using TypeScript, Fastify, Vercel AI, OpenAI/Gemini, RAG, vector search, and provider fallback logic; reduced report turnaround from days to about 15 minutes and enabled tens of thousands of citation-backed reports. • Implemented hybrid semantic/keyword retrieval across OpenAI vector stores, Elasticsearch, and Weaviate, improving source grounding, search reliability, and response quality across document-heavy workflows. • Developed backend APIs and data workflows with Fastify, Prisma, PostgreSQL, and REST endpoints supporting AI generation, search, analytics, and core product features. • Owned analytics instrumentation across Segment, Mixpanel, Customer.io, and HubSpot, including event taxonomy, validation, identify/group calls, CRM syncs, lifecycle events, dashboards, campaign triggers, and attribution. • Shipped Next.js/React features for commercial and government-facing applications used by hundreds of users, improving accessibility, usability, maintainability, and release confidence.
Infrastructure as Code team • Migrated applications to AWS, GCP, and Azure with Terraform/Terraform Cloud, reducing deployment time by about 25%. • Automated cloud infrastructure upgrades that saved 100+ work hours and improved security/consistency; wrote migration documentation and partnered with Agile teams to resolve blockers.
Enterprise Applications Team • Built React/Node.js internal tools with TypeScript, Docker, AWS, DynamoDB, Git, and Azure DevOps, improving workflow efficiency by about 30%. • Modernized cloud-backed workflows and deployment processes across internal applications, reducing deployment time by about 20%.