01 · Experience

Where I’ve worked

Five companies, fifteen years, one throughline: production systems where the failure cases are designed in, not discovered later.

Jun 2023 — Present
Remote · Eden Prairie, MN

Optum

Senior AI Full-Stack Engineer — Healthcare Platforms

  • Architect production AI-enabled healthcare services with retrieval, model calls, validation, business rules, and authorization separated into explicit components — so probabilistic model behavior never bypasses deterministic application controls.
  • Design LangChain/LangGraph agent orchestration where agents retrieve governed context, invoke approved tools, and request human approval for sensitive actions, with bounded, recoverable state transitions.
  • Build enterprise RAG pipelines — embeddings, FAISS, reranking, citation handling — over governed document sources, so responses stay tied to evidence.
  • Develop agent-evaluation harnesses with golden and adversarial cases and failure classification, in reproducible Docker environments with pinned dependencies and CI validation.
  • Ship human-in-the-loop React/TypeScript review interfaces and full-path observability with OpenTelemetry, Prometheus, and Grafana.
PythonFastAPILangChainLangGraphFAISSReactTypeScriptKafkaPostgreSQLRedisAWSKubernetes
Sep 2021 — May 2023
Remote

Palantir Technologies

Senior Software Engineer — AI & Cloud Platforms

  • Designed enterprise AI and data applications on Palantir Foundry — governed data models, Foundry Ontology, and operational workflows with lineage, permissions, and source-of-truth controls.
  • Engineered large-scale Spark/PySpark/SQL transformation pipelines handling schema evolution, data-quality validation, partitioning, and reproducibility.
  • Built ontology-backed React/TypeScript operational applications and FastAPI/REST services exposing governed data through stable, testable contracts.
  • Designed event-driven architecture with Kafka, PostgreSQL, and Redis — durable state, retries, duplicate handling, and explicit recovery behavior.
  • Produced platform standards, architecture guidance, and customer-facing walkthroughs; led architecture reviews, code reviews, and developer mentoring.
Palantir FoundryFoundry OntologyPythonPySparkSparkSQLFastAPIReactTypeScriptKafkaAWSDocker
May 2017 — Aug 2021
Remote

CVS Health / FirstBank

Senior Full-Stack Engineer

  • Built secure CVS Health services around pharmacy, benefits, eligibility, claims, and member workflows — authentication, authorization, data validation, and audit-aware logging across regulated boundaries.
  • Developed FirstBank transaction-oriented services treating duplicate requests, retries, partial failures, idempotency, and transaction boundaries as explicit design cases.
  • Implemented Kafka event-driven processing for healthcare and banking workloads with durable state and replay-safe handling.
  • Modernized legacy applications into independently deployable services with Docker, Git-based workflows, and CI/CD, backed by unit, integration, and regression test suites.
JavaSpring BootPythonReactREST APIsPostgreSQLSQL ServerKafkaRedisDocker
Sep 2014 — Mar 2017
Park City, UT

Backcountry

Full-Stack Engineer

  • Built customer-facing e-commerce across catalog, search, inventory, pricing, cart, checkout, and order workflows for high-traffic retail journeys.
  • Implemented checkout and order-processing integrations with explicit transaction boundaries, duplicate-request protection, and visible failure states across payment, inventory, and fulfillment systems.
  • Developed Node.js/Express services and React experiences backed by PostgreSQL, MySQL, and Redis, with asynchronous jobs designed to recover without duplicating shipments or notifications.
  • Refactored tightly coupled commerce logic into clearer service boundaries through incremental, test-backed changes.
JavaScriptNode.jsExpressReactPostgreSQLMySQLRedisREST APIs
Jul 2011 — Jun 2014
Boulder, CO

SendGrid

Backend Developer

  • Developed backend services for cloud email delivery infrastructure — stable public REST API contracts in front, durable distributed processing state behind.
  • Built high-volume queue and worker workflows with retries, duplicate protection, delayed execution, and failure classification.
  • Developed webhook and event-delivery pipelines with documented payload contracts and retry semantics external developers could depend on.
  • Implemented throttling, backpressure, and multi-tenant fairness controls so no single workload could destabilize shared capacity.
PythonREST APIsSQLRedisQueuesWebhooksDistributed systems