Chapter 01 · About
Senior Full-Stack
& AI/ML Engineer
15+ years in software engineering and 6+ years building production ML and applied AI systems, with staff-level ownership across architecture, hands-on development, evaluation, deployment, and production operations.
Hands-on with LLMs, RAG, embeddings, retrieval and ranking, agent workflows, Python/FastAPI, model serving, Kubernetes, and MLOps — improving GPU utilization ~35% and cutting retrieval latency 15%.
LLM & RAG
Systems
Built medical-record retrieval that grounds LLM applications at Novawave, moved customer-facing enterprise LLM applications from proof of concept into production at Larynx, and standardized ML delivery with MLflow at NextEra Energy.
LLM & RAG
Systems
Python
LLM & RAG systems
15+ yrs
Core Skills
Python · FastAPI · Java · TypeScript · Go · Node.js
LLMs · RAG · Embeddings · Reranking · pgvector
LangChain · LangGraph · Tool Calling · Agent Workflows
PyTorch · vLLM · Model Serving · Inference Optimization
React · Next.js · GraphQL · REST · WebSockets
AWS · GCP · Azure · Kubernetes · Terraform · Argo CD
MLflow · Kubeflow · LangSmith · Langfuse · Prometheus · Grafana
Production
ML
PyTorch, vLLM, model serving, and inference optimization. Feature engineering, anomaly detection, time-series forecasting, and recommendation and ranking systems for energy, logistics, and enterprise data.
Staff-Level
Ownership
Technical ownership from design through production operations.
Architecture & code reviews across application, retrieval, and model-provider layers
CI/CD & Kubernetes deployment automation for AI services
Observability with retries, timeouts, health checks, and SLO monitoring
Incident response and performance optimization for GPU workloads
Top Skills
Python, LLM and RAG engineering, FastAPI, PyTorch, and Kubernetes — 15+ years across healthcare AI, enterprise LLM applications, energy, logistics, fintech, and cloud platforms.
chapter 02:
experience
Chapter 02: Experience
Novawave LLC
Senior Applied AI/ML Engineer · Remote · Spokane, WA · Aug 2025 – Present
Larynx
Senior AI/ML Engineer · Remote · New York, NY · May 2023 – Jul 2025
Built customer-facing LLM features over enterprise data using embeddings, RAG, and prompt engineering for grounded generation
Designed evidence-constrained answer generation with context filtering and explicit insufficient-information handling — cutting the unsupported-answer rate 15%
Developed Python/FastAPI services and reusable retrieval and inference APIs, keeping AI service logic separate from product application code
Hardened prototypes with automated tests, containerized Kubernetes releases, scaling, and monitoring — reducing release-preparation time 20%
NextEra Energy
Lead ML Engineer · Remote · Juno Beach, FL · Jun 2021 – Apr 2023
Led development of ML services for energy asset analysis and optimization, turning operational telemetry requirements into deployable APIs
Designed Python pipelines connecting generation, consumption, and operational data with model inference — cutting pipeline runtime 20%
Implemented MLflow experiment tracking and model registries, reducing model-handoff time 25% with reproducible deployments
Owned deployment, model monitoring, and support while setting technical direction on architecture, reliability, testing, and CI/CD
Datex
Senior ML Engineer · Senior Software Engineer · Clearwater, FL · Dec 2017 – May 2021
Applied ML to warehouse and inventory telemetry; built Python pipelines connecting predictive workflows with WMS data, REST APIs, and databases — cutting processing time 15%
Automated recurring inventory and logistics processing, saving 4 hours of manual work per week
Delivered barcode-assisted pack verification across web UI, REST API, business logic, and database — reducing verification time 10%
Profiled and optimized SQL and legacy verification paths, reducing query execution time 20% with regression coverage
Earlier
Career
FIS · Microsoft · Google · Jacksonville, FL · Redmond, WA · Kirkland, WA · 2010 – 2017
FIS · Senior Full-Stack & Backend Engineer (2015 – 2017) — Java payment services for the Open Payment Framework, reducing p95 latency 15%
Microsoft · Full Stack & Backend Engineer (2011 – 2015) — Azure Preview Portal features and Windows Azure Java integrations, cutting workflow load time 15%
Google · Software Engineering Intern (2010) — Gmail calling and Google Voice integration backend
chapter 03:
skills
Chapter 03 · Tools
Tech
Stack
Python, FastAPI, PyTorch, vLLM, LangChain, LangGraph, pgvector,
React, Next.js, AWS, GCP, Azure, Kubernetes, Terraform, MLflow, and full-stack production delivery.
Languages & Tools
LanguagesPython · Java · TypeScript · JavaScript · Go · SQL · Bash
AI / ML & GenAILLMs · RAG · Embeddings · Semantic Search · Reranking · PyTorch · vLLM · LangChain · LangGraph · Tool Calling · Agent Workflows · Model Evaluation
Backend & FrontendFastAPI · Node.js · Express · GraphQL · REST · Kafka · WebSockets · React · Next.js · Vue · Angular
DataPostgreSQL · pgvector · Redis · Spark · PySpark · Airflow · BigQuery · ETL/ELT · Feature Pipelines
Cloud & MLOpsAWS · GCP · Azure · SageMaker · Bedrock · Vertex AI · MLflow · Kubeflow · Docker · Kubernetes · Terraform · Argo CD
Evaluation & ObservabilityLangSmith · Langfuse · DeepEval · A/B Testing · Prometheus · Grafana · OpenTelemetry · Tracing · SLO Monitoring
Impact 1 2 3 4
1
2
3
4
~35% better GPU utilization · 20% fewer duplicate model calls · 15% lower retrieval latency · 15% fewer unsupported answers · 25% faster model handoff · 20% faster releases.
Chapter 03 · Education
Education
University of Washington — Master's and Bachelor of Science in Computer Science.
University of Washington
Master's Degree, Computer Science
Sep 2009 – Jun 2011
University of Washington
Bachelor of Science, Computer Science
Sep 2005 – Jun 2009
Always
Learning
Continuously evaluating new LLM, retrieval, and inference tooling — from agent frameworks and evaluation harnesses to GPU serving optimizations — and bringing what works into production.
Design
AI Systems
AI Systems
Architect retrieval, inference, and agent workflows: embedding-based retrieval, ranking and reranking, grounded generation, and structured outputs with explicit handling of insufficient evidence.
Build
Production Services
Production Services
Ship Python/FastAPI services, reusable retrieval and inference APIs, and React features that connect product workflows to LLMs without duplicating service logic.
Operate at
Scale
Scale
Own Kubernetes deployments, CI/CD, observability, and incident response. Tune GPU utilization, latency, and reliability for AI workloads in production.
