Aaron Johnson

Portland, OR · LinkedIn · GitHub · amj@mqqn.net

---

Data engineer and performance specialist. I build data pipelines, optimize code, and make systems scale. Background in applied mathematics and distributed computing. Currently exploring AI-assisted development and GPU kernel optimization.

---

Recent Work

GPU Kernel Optimization (2026)

• Triton kernel development for parallel prefix sum algorithms
• Systematic performance tuning on NVIDIA A10G
• Published findings on AI safety evaluation systems

AI-Assisted Development (2024-present)

• Building apps with Claude Code: Step Quest, Digital Forester
• Developing patterns for Claude skills and project-specific AI context
• Documenting workflows for technical and scientific applications

---

Experience

ZeroTier · Remote · 2023-2025

• Senior Data Engineer / Solutions Architect
• Business metrics reporting across Stripe Sigma, BigQuery, PostgreSQL
• Data pipelines with Clickhouse and Kafka for high-volume event processing
• Technical support for enterprise sales using Claude Code for codebase exploration

Intel · Hillsboro, OR · 2018-2023

• Math Algorithm Engineer
• Performance engineering for mathematical computation libraries
• Intrinsics programming for matrix operations on embedded platforms
• Eigen vs oneMKL performance analysis
• Kubernetes, OpenStack, bare-metal infrastructure management

Jama Software · Portland, OR · 2015-2016

• DevOps Engineer
• AWS high-availability deployments with Beanstalk and Docker
• Python/Boto CloudFormation automation
• Scrum master for DevOps team

Dark Horse Comics · Milwaukie, OR · 2010-2013

• Senior Software Developer
• Python/Django digital comic delivery platform
• Scaled architecture for unpredictable traffic spikes
• PostgreSQL, CloudFront, S3, RabbitMQ, Puppet

Earlier: ESRI, Metal Toad, Cambia, Peak Hosting, RightScale, Patterson Dental, Flint Hills Scientific

---

Technical Skills

Area	Technologies
Languages	Python, SQL, C, Ruby, Shell
Data	PostgreSQL, BigQuery, Clickhouse, Kafka, Redis
Cloud	AWS (EC2, S3, Lambda, ECS), Docker, Kubernetes
GPU	Triton, CUDA concepts, performance profiling
Infrastructure	Linux (20+ years), OpenStack, bare-metal
AI Tools	Claude Code, prompt engineering, AI-assisted development

---

Education

M.A., Mathematics · University of Kansas

Distributed parallel computing (PVM, MPI), numerical analysis, scientific visualization. High-performance platforms: SGI, HP Convex, DEC Alpha Beowulf cluster.

---

Links

• Full profile: LinkedIn
• Code: GitHub
• Writing: aaronjohnson.github.io
• Apps: mqqn.net

---

Last updated: February 2026