- Joined Cognition
- Jul 2025
- How they joined
- With Windsurf
- Before Windsurf
- Volley AutomationSoftware Infrastructure Engineer
- Status
- Current
- Cognition signed a definitive agreement to acquire Windsurf (formerly Codeium) in July 2025, days after Google DeepMind hired Windsurf's CEO and research leaders; Vanderwerf was on Windsurf's staff at the time.Jul 14, 2025
- Member of Technical Staff at Cognition (maker of Devin and, since July 2025, Windsurf) from July 2025; the start date coincides with Cognition's acquisition of Windsurf.Jul 2025
- On Windsurf launch day he posted on LinkedIn: "Today we are launching our own IDE, Windsurf, powered by a new way of working with AI, which we have called Cascade. I use Cascade all the time to make changes across all our production code, whether it's Kubernetes resources, helm charts, Terraform, or any of the supporting tooling and scripts. Here's a quick example of Cascade making a script to check Kubernetes pods near arbitrary resource limits across a cluster. Check out Windsurf for yourself at windsurf.ai"Nov 13, 2024Today we are launching our own IDE, Windsurf, powered by a new way of working with AI, which we have called Cascade. I use Cascade all the time to make changes across all our production code, whether it's Kubernetes resources, helm charts, Terraform, or any of the supporting tooling and scripts. Here's a quick example of Cascade making a script to check Kubernetes pods near arbitrary resource limits across a cluster. Check out Windsurf for yourself at windsurf.ai
- Joined Codeium/Windsurf as a Production Engineer in September 2024, about two months before the Windsurf IDE launch.Sep 2024
- His launch post shows that by November 2024 his day-to-day at Codeium covered production infrastructure (Kubernetes resources, Helm charts, Terraform and supporting scripts), and that he dogfooded Cascade, the agentic feature, on that infra code.Nov 13, 2024Today we are launching our own IDE, Windsurf, powered by a new way of working with AI, which we have called Cascade. I use Cascade all the time to make changes across all our production code, whether it's Kubernetes resources, helm charts, Terraform, or any of the supporting tooling and scripts. Here's a quick example of Cascade making a script to check Kubernetes pods near arbitrary resource limits across a cluster. Check out Windsurf for yourself at windsurf.ai
- Robotics Reliability Engineer at Nuro (Mar 2020 - Nov 2022): he owned, developed and maintained the Secure Bot Network, which provided HA secure VPN connections to Nuro's whole fleet (Prii and robots) for remote monitoring, telemetry log upload, disk-decryption key distribution and mTLS gRPC remote command.Mar 2020
- Software Infrastructure Engineer at Volley Automation (Sep 2023 - Aug 2024), where he re-architected the on-prem central compute cluster onto OnLogic rugged industrial servers, provisioned via Ansible and Terraform.Sep 2023
- At Nuro he helped bring up the P2/R3 platform and built a hot-plug external-modem load-balancing fix, which let the platform test on open roads while waiting on vendor firmware patches.2022
- Public headline says "Member of Technical Staff at Cognition".
- At Volley Automation he built AWS infrastructure for a 24/7 livestream of the company's production garage in downtown San Francisco and wrote Ansible/Kubernetes code to configure its autonomous EV charging infrastructure.2024
- Software Engineer on Spreetail's DevOps Chapter (Jun 2019 - Mar 2020), working with Elasticsearch, Packer, Terraform, Prometheus, HAProxy/Nginx and Okta/Vouch-Proxy SSO. Before that he was a DevOps intern there (May 2018 - May 2019) on Docker, Kubernetes, Istio, Helm, Chef and GitLab-CI.Jun 2019
- B.S. in Software Engineering from Iowa State University (2015-2019), magna cum laude (3.75/4.0), with a minor in Cyber Security.2019
● Fully re-architected on-prem central compute cluster from Dell PowerEdge Server to OnLogic rugged industrial server, fully provisioned via Ansible & Terraform, to better support business proposition & sales as well as reduce complexity & improve development velocity ● Architected all infrastructure in AWS for 24/7 livestream of production garage in downtown SF for sales initiatives ● Updated legacy Terraform code (~ 9 months out of sync) to reflect current garage deployment state including: AGVs, bay compute, and central cluster server infrastructure. ● Developed deployment pipelines for automated garage deployments using BitBucket, AWS CodeBuild, and a combination of Bash and Python scripting (with multithreading to enable simultaneous Terraform deployments to each piece of infrastructure). ● Instrumented Prometheus monitoring of all workloads running in Kubernetes. Configured AlertManager reporting via Slack with 70 alerting rules. These alerts found previously unknown bugs in both the VMWare csi-driver and FluentBit, which I later patched. ● Created security scan reports of all Docker images running in production & staging environments with rankings according to CVE severity. ● Integrated CodeCov scans into PR pipelines for all production code including: C++, C#, & Python. Reports of coverage diffs are posted to all new PRs going into the main branch. ● Developed all Ansible code to configure autonomous EV charging infrastructure as well as K8s manifests, managed via Terraform, for deploying all relevant software and workloads to said infrastructure.
● Owned, developed, and maintained the Secure Bot Network. Responsible for HA secure VPN connections to all of Nuro’s fleet (Prii, Robots, etc.) for: debugging, remote monitoring, lightweight live log upload of security/telemetry data, secure disk decryption key distribution, secure remote gRPC through mTLS for offboard remote command control, etc. ● Assisted in the bring up of Nuro’s P2/R3 platform, debugged the issues observed, and appropriately triaged all issues to the relevant owning teams to ensure proper long-term resolution. ● Developed a solution to enable hot-plugging an external modem to the onboard compute & automatically add it into the load balancing. Which enabled testing the P2/R3 platform on open roads while awaiting firmware patches from the third party modem provider. ● Debugged all kinds of issues live on Nuro’s fleet, RCA-ed to appropriate HW/SW components/modules and triaged to the appropriate owner to ensure proper long-term resolution. ● Led the transition of RRE infrastructure to GKE by developing relevant Terraform modules. ● Led the transition of log upload from old in house tools to FileBeat & Logstash with an output sink to Sumologic for greater scalability & automated security alerting on security critical logs & metrics. ● Architected a new secure solution for engineers to SSH into live vehicles for debugging purposes, sunsetting a single server that had been in use for years while considerably increasing observability and auditability.
DevOps Chapter Additional Technologies: Data Clusters: Elasticsearch Infrastructure As Code: Packer, Terraform Monitoring: Prometheus Proxies / Load Balancers: HAProxy, Nginx SSO Authentication: Active Directory, Okta, Vouch-Proxy
DevOps - Docker, Kubernetes, Istio, Helm, Chef (Ruby), Octopus Deploy, GitLab-CI, Consul Cloud Platforms - Azure Scripting Languages - Powershell, Bash Backend Languages - C# Libraries - Hangfire, Topshelf, NewRelic Hypervisors - Hyper-V Full time in the summer. Continuing remote part-time work throughout the semester.
Computer Science 362 - Object-Oriented Analysis and Design (Mr. Robert Ward)
Software Engineering 339 - Software Architecture and Design (Dr. Lotfi Ben-Othmane)
Magna Cum Laude - 3.75/4.0 Minor in Cyber Security
Today we are launching our own IDE, Windsurf, powered by a new way of working with AI, which we have called Cascade. I use Cascade all the time to make changes across all our production code, whether it's Kubernetes resources, helm charts, Terraform, or any of the supporting tooling and scripts. Here's a quick example of Cascade making a script to check Kubernetes pods near arbitrary resource limits across a cluster. Check out Windsurf for yourself at windsurf.ai
