Brian Waldon

Diamanti

VP, Product • Dec 2018 — Present

Diamanti was founded on the premise that containers and hardware offload technologies could be combined to operate incredibly efficient infrastructure. I joined this team of highly-skilled engineers to expand the capabilities of the Diamanti platform into the cloud.

Some other things I have accomplished while at Diamanti:

• Developed and assisted in implementation of Diamanti Spektra & Ultima roadmaps, which strategically expanded the platform into full-software and cloud-based environments while introducing new multi-cloud and multi-cluster capabilities
• Designed, implemented and operated Diamanti Central, which directly interfaced with Diamanti customers via a web application to distribute product licenses, collect platform telemetry, and serve hosted sandbox environments
• Led a team of product managers and UX/visual designers to provide direction for the engineering teams
• Acted as chief product tester to ensure high quality and accurate feature development
• Represented Diamanti publicly and privately: customer briefings and product testing, analyst briefings, press engagements, etc

You may be asking yourself, "why would Brian switch from engineering to product?" To to put it simply, Diamanti needed me to focus on product strategy and roadmap more than write software on day one. My technical background was vital in my success as a product leader, since it helped me come into a new organization with empathy for the challenges of shifting from a hardware to a software/cloud mindset.

Planet Labs

Principal Engineer • Jan 2016 — Aug 2018

Planet has done some absolutely amazing things. I was able to join after the initial constellations of earth imaging satellites were in orbit, and I could build on the immense pipeline of data already being generated. My major focus during my time at Planet was to bolster the existing public-facing imagery services, and develop new internal and external services that could derive analytics from imagery using machine learning.

Some of my notable achievements include:

• Architected and implemented a large-scale analytics platform that empowered the data science team to effectively process live imagery, and serve the derived insights back to public customers.
• Deployed and operated Kubernetes as primary infrastructure platform for public-facing serviecs on AWS and GCP
• Developed new means of delivering imagery to customers, directly via API and asynchronously through the cloud
• Led operationally-focused teams of DevOps engineers and SREs towards a more sane existence (i.e. on-call rotations quieter, less frequent, and overall higher service uptime)

CoreOS

Engineering Team Lead • Nov 2013 — Dec 2015

I defined and implemented the initial architecture of Tectonic, CoreOS' flagship commercial product. Tectonic delivered Kubernetes along with a custom management plane to enterprise customers. I was instrumental in establishing the early access program, and I built a cohesive development team around the product, eventually releasing GA in early November, 2015.

My role as a team lead was to ensure that active development on the product continued to move forward efficiently. My primary responsibilities included:

• Evolve the overall architecture and roadmap of the product, engaging internal stakeholders and external customers
• Assist with company goal-setting and performance tracking via the OKR process
• Drive day to day operations of a team of 5+ engineers, ensuring proper software development practices are observed and happiness is abundant
• Develop Tectonic codebase alongside team members, providing guidance through code review and technical discussion
• Manage releases to customers, generation of documentation and clean migration paths to preserve backwards compatibility
• Ensure customer success via proactive conversations and reactive L3 support

During the development of Tectonic, I pioneered two major open-source projects, dex and kube-aws, both of which are detailed below under Projects.

Prior to CoreOS' decision to build Tectonic, I was responsible for CoreOS' custom cluster management solution, fleet. The project provides a cluster-level abstraction atop systemd, allowing users to safely operate services in an ephemeral environment. More information is available under Projects.

Nebula

Architect & Software Development Manager • July 2012 — Nov 2013

Nebula aimed to bring OpenStack to the datacenter in the form of a hardware appliance, the Nebula One Cloud Controller. I joined Nebula to lead an engineering team responsible for all OpenStack-focused development. I also contributed as a software engineer and architect.

OpenStack

Project Technical Lead • Oct 2011 — Sep 2012

I was elected by the OpenStack community to serve as the Project Technical Lead (PTL) of Glance for two consecutive release cycles. Glance acts as the central image service. As the PTL, I was responsible for leading project development and engaging with the OpenStack community.:

Rackspace

Software Developer, Rackspace Cloud • Jan 2011 — Apr 2012

Worked alongside public and private cloud teams to deliver OpenStack to customers.

Software Developer, Rackspace Email & Apps • Jan 2008 — Apr 2011

Built backend systems in Python & PHP related to email routing and spam filtering.

fleet

fleet is a cluster-level abstraction of systemd, enabling users to safely operate services in an ephemeral environment while incurring minimal resource and cognitive overhead.

• Provides schema-driven REST-like API for users to interact with the system
• Interacts with systemd via DBus APIs
• Uses etcd for distributed locking & persistent data storage
• Operated production CoreOS services for ~2 years
• Deployed by external companies to coordinate thousands of hosts

dex

dex is a federated identity management service. The primary goal of dex is to provide a powerful, flexible, secure solution to identity management while minimizing overhead on both end users and deployers of the system.

• Frontend communication via OpenID Connect, similar to OAuth2
• Users self-authenticate through federation of control to arbitrary backends (Google, Facebook, LinkedIn, GitHub, LDAP, local DB, etc)
• Signing keys automatically rotated, sensitive data encrypted at rest

Smaller Notable Projects

In addition to the larger projects above, I have also been responsible for many smaller projects over the years:

coreos-cloudinit

• Inspired by Ubuntu's cloudinit tool
• Used to configure CoreOS Linux hosts via YAML
• Aware of metadata services on most major cloud providers: AWS, GCP, Rackspace, Digital Ocean, etc
• Managing hundreds of thousands of active CoreOS Linux hosts
• Shipped directly in CoreOS Linux image

go-etcd

• Completely rewrote original client bindings based on what was learned from development of fleet
• Official golang bindings for etcd APIs
• Used by by fleet for locking, data persistence, liveness checks

go-systemd

• Systemd unit file parser
• Reverse-engineered systemd source code
• Package written for fleet to (de)serialize unit files