Aaron Cupp

Sacramento, CA · Remote

Senior / Staff Infrastructure · SRE · Platform Engineer. Open to senior IC and engineering-leadership roles — infrastructure, SRE, platform. Remote-first.

Summary

Infrastructure and site reliability engineer with over two decades building and operating cloud and on-premises platforms across GCP, AWS, and Azure. Deep hands-on work in Kubernetes, GitOps, and Infrastructure as Code (Terraform / OpenTofu), plus infrastructure control evidence for SOX, SOC 2, and PCI-DSS under Big Four external audit. Has built and run a multi-cloud SRE function to 99.98% uptime with 365-day coverage — and stayed on the keyboard through it. Builds the platform plumbing other teams ship on, and has a habit of finding the failure nobody was looking for before it reaches production. Stays sharp through active open-source work in Rust, Go, and Kubernetes tooling — shipping tested, CI-backed projects end to end.

Experience

Senior Infrastructure Engineer, Commerce (Feedonomics) — Remote, Jan 2025 – Present

Sole infrastructure IC for FDX Infrastructure — the platform behind Commerce’s product-data feed and marketplace optimization engine. Scope spans platform reliability, infrastructure as code, audit and compliance evidence, and internal tooling.

Built the GCP Terraform foundation greenfield — the infrastructure-as-code layout and cloud configuration other teams now build on, including an in-progress colo-to-cloud migration. Among the top contributors by commit volume.
Own infrastructure control evidence for SOX, SOC 2, and PCI-DSS under Big Four external audit — evidence packages, control-gap remediation, and forensic timeline reconstruction from recovered system state.
Caught an exhausted production disk quota that would have failed six pending deployments; reworked the affected changes same-day and built tooling so the condition surfaces instead of failing silently.
Operate Kubernetes with Flux GitOps and a unified observability stack — Grafana, Prometheus, Mimir, Loki, Alloy, AlertManager — consolidating metrics and logs from many clusters into a single pane of glass.
Author Terraform / OpenTofu modules and Atlantis automation, keeping environments reproducible and reviewable in pull requests.
Wrote two operational procedures now used by the team: in-place rescue-mode instance repair and GCE emergency-mode fstab recovery.
Lead incident response and the on-call rotation; mentor engineers on Linux fundamentals, GitOps practice, and production standards.

Founder & Principal Engineer, Tech-Noid Systems — Sacramento, CA, 2008 – Present (full-time Aug 2023 – Jan 2025)

Self-directed technical venture founded in 2008 as a self-built online radio platform, run full-time between engineering roles and ongoing today. Now a bass music collective with its own soundsystem, self-hosted infrastructure, open-source tooling, and live performance technology.

Designed and operate a self-hosted, multi-node Kubernetes cluster (mixed x86_64 / arm64) with Flux GitOps, MetalLB, cert-manager, Helm, and full observability.
Authored obs-radio-output — a native OBS Studio plugin in C/C++ against libobs that streams audio directly to Icecast / SHOUTcast, with macOS universal, Linux, and Windows builds and a signed, notarized macOS package.
Manage infrastructure as code with Terraform / OpenTofu, Terragrunt, and Ansible, including Discord-as-Code modules; rebuilt the web platform on Hugo.
Built out a QSC-based soundsystem and run live bass-music events on it; the collective’s streams, radio feed, and shows all run on infrastructure I operate.

Manager, Site Reliability Engineering, Sight Machine — San Francisco, CA (Remote), Nov 2021 – Aug 2023

Built and led the SRE function — grew the team to three full-time engineers plus two offshore teams for 365-day coverage, hiring, onboarding, and developing engineers into an on-call-capable organization.
Engineered multi-cloud, Kubernetes-based infrastructure for a big-data SaaS platform across GCP, Azure, and AWS using Python, Kafka, and Terraform, sustaining uptime above 99.98%.
Drove a 10% reduction in monthly server spend through autoscaling and spot-instance systems; earned the Golden Gear award for cost measures saving $5k monthly.
Led the Infrastructure Modernization Project, upgrading to modern Kubernetes configurations and increasing security and automation across the stack.
Standardized onboarding, on-call rotations, incident response, and change management. Member of the Company Security Response Team and the Engineering Technology Leads Team.

Senior Site Reliability Engineer, Sight Machine — Remote, Aug 2021 – Nov 2021

Elevated to Lead Senior SRE within two months. Pioneered Terraform with Atlantis/GitHub CI/CD and established a GitOps framework for all cloud infrastructure.

Founder, Emotional Support Pizza — Sacramento, CA, 2019 – 2021

Woodfired ancient-grain sourdough concept built during a break from tech. Built the full digital presence — website, brand, and online-ordering flow — which reignited hands-on engineering and led directly back into infrastructure work.

DevOps Engineer, JUUL Labs — San Francisco, CA (Remote), Jul 2017 – May 2019

Delivered 24/7 on-call support for all e-commerce systems; primary support for Windows systems running the Acumatica ERP.
Implemented log-shipping-based clustering for SQL Server Enterprise, supporting business expansion into multiple regions.
Developed infrastructure as code with Terraform modules; migrated legacy e-commerce systems from PAX Labs to JUUL Labs after the company split.

Senior Infrastructure Engineer, PAX Labs — San Francisco, CA (Hybrid), Nov 2016 – Jul 2017

Led expansion of AWS infrastructure from 4 to over 100 systems with autoscaling policies for high-traffic volumes.
Standardized system deployment using Jenkins and Fabric, and implemented company-wide automated paging via PagerDuty.

Systems Engineer, study.com — Mountain View, CA, Sep 2014 – Sep 2016

Led migration of cloud infrastructure from Rackspace to AWS, growing the fleet from 10 to over 300 systems.
Implemented full server lifecycle management with Packer, Puppet, PuppetDB, and Foreman, integrated with Jenkins and Rundeck.
Developed and enforced a disaster-recovery plan safeguarding the data-warehouse project. Mentored the junior systems engineers on the team.

Senior Linux Administrator, Meta Interfaces — San Francisco, CA (Hybrid), Oct 2012 – Sep 2014

Administered 1,000+ physical Linux servers and network hardware (Cisco, F5 BIG-IP) for a data center supporting high-demand streaming video.
Led the initial AWS proof-of-concept that set the groundwork for cloud strategy, and replaced physical servers with a KVM / libvirt virtualized environment.
Implemented monitoring with PagerDuty and Nagios.

Selected projects & open source

Audiophore (Rust) — engine bridging VJ audio analysis (Synesthesia OSC) to real-time lighting across Hue, WLED, DMX / Art-Net, sACN / E1.31, and ILDA lasers. Sub-30 ms latency, Lua-scripted show files, Tauri + Svelte native app.
obs-radio-output (C/C++) — native OBS Studio plugin streaming straight to Icecast / SHOUTcast. Public beta on macOS, Linux, and Windows.
annoybots (Go) — a single Go binary driving many IRC / Twitch / Discord networks at once over a shared Redis bus, with a Markov engine and sandboxed Lua plugins. Race-tested CI, multi-arch distroless images, GitOps (Flux) deployment on Kubernetes.
claude-project-kit — open-source workflow kit that keeps AI-assisted coding grounded across repos: durable project memory, phased plans, and lifecycle commands, backed by a Bats test suite, CI, and ADRs. Adopted for personal and day-job delivery.
The bench stack (Python) — benchhud, bench-parts, and benchhud-intake: an OpenCV / v4l2loopback compositor fusing microscope, thermal-imaging, and instrument telemetry into a single live video feed, backed by a separate self-hosted inventory service whose REST API the HUD and an intake service both consume. Runs on the cluster.
3D Printer Models and Clickfinity (OpenSCAD) — thirty parametric models with CI-validated meshes and per-model releases, plus a clean-room magnet-free Gridfinity baseplate generator.
recipe-card-maker (Python / SvelteKit) — FastAPI backend, SvelteKit frontend, LLM-based recipe extraction feeding a Markdown-to-PDF pipeline. Containerized with CI and a full test suite.
Snapmaker U1 tooling (Shell / Swift) — binary analysis and patching scripts for vendor firmware images, RFID diagnostics, and an iOS app that writes OpenSpool NFC tags.

Technical skills

Cloud & orchestration: GCP · AWS · Azure · Kubernetes · AKS · Flux CD · GitOps · Helm · Docker · containerd · MetalLB · cert-manager · linkerd
Infrastructure as code: Terraform · OpenTofu · Terragrunt · Ansible · Puppet · Packer · Atlantis
Observability & reliability: Grafana · Prometheus · Mimir · Loki · Alloy (OpenTelemetry) · AlertManager · incident response · on-call · blameless post-mortems
CI/CD & tooling: GitHub Actions · Jenkins · CircleCI · Rundeck · Fabric · Vault · HAProxy · nginx · Kafka
Compliance: SOX · SOC 2 · PCI-DSS control evidence · external audit support · change management
AI-assisted delivery: Claude Code · agentic development workflow design · durable project context and memory systems · AI-assisted IaC authoring and review
Languages: Python · Go · Rust · Bash · C/C++ · PHP · JavaScript
Platforms & data: Linux (Ubuntu, RedHat, CentOS) · Windows Server · KVM / libvirt · PostgreSQL · MySQL · SQL Server · MongoDB · Redis

Earlier experience

Computer Technician II — Elk Grove Unified School District, Elk Grove, CA — 2005 – 2012
Developer & Systems Administrator — Groundworks Environmental, Folsom, CA — 2004
Intern, Network Technologies — Stanislaus County Office of Education, Modesto, CA — 2003 – 2004
Student Employee, Information Systems — Yosemite Community College District, Modesto, CA — 1999 – 2003

Education

A.S., Computer Science — Modesto Junior College, Modesto, CA — 2004
Computer Programming Specialist Certificate — Modesto Junior College, Modesto, CA — 2003
AWS Operations Bootcamp — AWS Summit, San Francisco, CA — 2013