Skip to main content

Resume

Joshua Powers · Senior Engineering Manager at Chainguard

I build and lead remote engineering teams that maintain widely used open source infrastructure, including Ubuntu Server, cloud-init, Telegraf, and Wolfi. I’ve kept shipping code the whole time.

I’ve moved between writing software and leading the people who write it four times, usually on the same project. That’s how I lead infrastructure work best, because you can’t set a roadmap for a codebase you’ve stopped reading. I automate the toil, measure what matters, and teach in public.

17 years engineering · leading since 2018 · teams from 3 to 18 · maintainer of Telegraf, cloud-init, and Wolfi · 95+ technical posts

What I do

Build and lead remote teams

I’ve led teams of many sizes. At InfluxData I led three engineers maintaining Telegraf, where the manager also had to be a committer. At Chainguard I lead eight engineers on Sustaining Automation, growing to eleven, and I led eighteen across the wider Sustaining organization at its largest. At Canonical I led the worldwide Ubuntu Server team, which was distributed years before remote work was common.

I’ve run sprints and planning cycles in Denver, Budapest, New York, and Seattle, and turned twice-yearly time together into roadmaps that distributed teams actually delivered. I’ve hired, onboarded, and promoted engineers into senior and maintainer roles, and I’ve been the person they page.

Maintain open source projects

For three years I maintained Telegraf, the open source metrics collection agent, while managing the team that owned it. I handled releases, code review, the plugin ecosystem, and the community. Along the way we shipped Apple Silicon and RISC-V support and a rewritten, faster line protocol parser.

At Canonical I was a core contributor and quality owner for cloud-init, which runs at first boot on nearly every instance on every major cloud. I built its CI pipelines and integration test framework from scratch and led its move from Launchpad to GitHub. I also organized the annual cloud-init summit, which brought Amazon, Microsoft, Google, Cisco, Oracle, Red Hat, and SUSE into one room.

Today I lead the team that maintains Wolfi, the container-native Linux (un)distribution behind Chainguard Images.

I’ve built communities around the code too. Daily community triage cut Telegraf’s open issues by 75%. I’ve written blog posts for nine years, produced instructional videos and classroom training, and earned official Ubuntu Membership for sustained contribution.

Secure the software supply chain

Chainguard sells security: hardened, minimal, continuously patched container images. I lead the automation behind CVE remediation, package updates, and image releases at scale.

Before Chainguard, I created cloud-init’s security disclosure process, which defines how issues are reported, evaluated, fixed, and disclosed. I automated delivery of Ubuntu cloud images to Google Cloud, including Common Criteria and FIPS hardening, and advised partners on security practices. At InfluxData I added fuzz testing and automated security notifications to Telegraf’s release pipeline. My first security work was at HP, where I built a hardened Debian-based operating system adopted across product lines.

Run engineering on metrics

I own remediation SLAs and built the systems that measure them for each customer: one day for KEV vulnerabilities, seven or fourteen days for CVEs depending on severity, and fourteen days for package updates.

I use data to remove process as often as to add it. After analyzing years of Telegraf release-candidate feedback, I showed that the release candidates produced almost no useful signal, and I replaced them with nightly builds and per-PR test artifacts that the community actually used. At Canonical I developed the metrics methodology that leadership used to decide where to invest in Ubuntu Server.

Metrics are also a subject I know from the inside: I spent three years maintaining an open source agent built to collect them.

Develop with agents

I’m leading agentic workflow adoption on my team at Chainguard to speed up CVE fixes and grow output faster than headcount. That has meant writing agent skills for our core workflows, such as new package development and package migrations. The skills turn tribal maintainer knowledge into repeatable processes that an agent can run.

A decade of automating triage, CI, fuzzing, and release engineering led up to this. The difference now is that the agents write the patches too.

Experience

Chainguard

  • Senior Engineering Manager, 2025–present
  • Engineering Manager, 2024–2025

Lead the Sustaining Automation team, which builds the automation behind CVE remediation, package updates, and image releases, and maintains Wolfi.

InfluxData

  • Engineering Manager, 2023–2024
  • Senior Software Engineer, 2021–2023

Maintained Telegraf, then led the fully remote team that owns it.

Canonical

  • Engineering Manager, Public Cloud, 2020–2021
  • Engineering Manager, Ubuntu Server, 2018–2020
  • Senior Software Engineer, 2016–2018

Built cloud-init’s CI and test framework, led the worldwide Ubuntu Server team, then led delivery of Ubuntu images to Google Cloud.

Hewlett Packard

  • Senior Software Engineer, 2014–2016
  • Software Engineer, 2009–2014
  • Software Engineering Intern, 2007–2009

Built a hardened Debian-based operating system used across HP products, and performance tests for enterprise servers.

Education

Georgia Institute of Technology

  • M.Sc. Computer Science, Interactive Intelligence, 2014–2016

Colorado State University

  • B.A. Economics, 2005–2009
  • B.S. Applied Computing Technology, 2005–2009

Skills

Go (Golang) · Python · Bash · Linux · Debian and Ubuntu · Kubernetes · Docker and OCI containers · Google Cloud · AWS · Azure · Telegraf · InfluxDB · Prometheus · Grafana · CVE remediation · software supply chain security · SBOM · FIPS · CI/CD · GitHub Actions · release engineering · agentic AI workflows