Experience
Staff-level infrastructure engineer in San Francisco. Kubernetes, Terraform, GitOps and ML platforms on AWS, Google Cloud and Azure, and lately a lot of GPU inference.
zac@zacp.xyz·linkedin.com/in/zacpete·San Francisco, CA
Roles
Healthcare company · Staff DevOps Engineer, AI/ML infrastructure
2025–present- Built multi-cloud Kubernetes infrastructure with Terraform and GitOps delivery.
- Developed secure ML and FHIR data workflows on AWS, Postgres and Snowflake.
- Connected MLflow and Airflow into the team's MLOps platform.
Global energy company · Staff DevOps Architect, AI/ML infrastructure
2024- Cut infrastructure costs by 40% with a GCP and Kubernetes redesign.
- Took model deployment from weeks to hours with Kubernetes, Kubeflow and MLflow.
- Brought SRE practices and observability to production systems.
Consumer apparel company · Senior Cloud Architect
2023–2024- Cut cloud spend by 60% through GCP architecture and automated Terraform provisioning.
- Built multi-cloud Kubernetes foundations for containerized applications.
- Moved JavaScript and Node.js delivery onto Argo CD.
Digital marketing company · Senior DevOps Engineer
2021–2022- Senior individual contributor on platform infrastructure.
- Tied application deployment to cloud-native DevOps practices.
Cloud consultancy · Managed Services Practice Lead (joined as Cloud Engineer)
2019–2021- Designed, built and ran cloud infrastructure for managed-services clients.
- Led the SRE and DevOps team for the managed-services division.
- Worked across AWS, Azure and Google Cloud.
Projects
Running GLM-5.1 on a rented 8× H200 node
One node, one model, one user. The setup that made it repeatable.
A Kubernetes operator for coding-agent benchmarks
CRDs for profiles and runs, a scheduler that checks headroom, and a database that remembers every attempt.
A five-node k3s build fleet for CI
Build where the cache already is, and never run the same validation twice.
Skills
- Cloud
- AWS, Google Cloud, Azure
- Kubernetes
- GKE, EKS, AKS, k3s, Helm, Argo CD, Flux
- Infrastructure as code
- Terraform
- ML platforms
- MLflow, Airflow, Kubeflow
- Inference
- vLLM, SGLang, speculative decoding, FP8
- Observability
- Prometheus, Grafana, Loki, OpenTelemetry
- Build and delivery
- BuildKit, Harbor, GitLab CI, Woodpecker CI
- Languages and data
- Go, Python, TypeScript and Node.js, PostgreSQL, Snowflake, Kafka
Education
B.S. in Computer Science, California State University, Fullerton.
Get in touch
If something here sounds like your problem, send me a few lines about what you're running.