FotoLog was born out of frustrations trying to share trip activity with friends & family. It helps me curate, map, and share photographs (fotos) with others. It is not a general purpose photo backup & management application. For now, it is for personal use but designed it to be useful for site management, so that groups such as parks & forestry departments or property managers could track issues that need attention.

LogBus was developed to address some of the shortcomings of other log shippers at the time:

• inability to manage state which is needed in order to aggregate/roll-up
• unnecessarily complicated DSL-like configuration instead of trusting users with basic programming language primitives
• inability to split the event stream and rejoin further down the processing pipeline

Technology that I have production experience with:

• LANG: TypeScript, JavaScript, Python, Java, PHP
• INFRA: Docker, Kubernetes, Bash, Salt, Puppet, iPXE, Kickstart Jenkins, Rundeck
• DATA: Postgres, MySQL, Redis, Elasticsearch, Kafka
• SYSTEMS: Linux (Debian, Red Hat), Cloud (AWS, DigitalOcean), Xen
• WEB: React, NGINX, Apache, HTTP, DNS, SSL/PKI

Upgraded their minimum viable product (MVP) to a proper web application that could handle much larger scale and enable developers to debug & add features using common tool & best best practices. The broad areas where the greatest impact was felt:

• More mature change control: used GitLab to build well-defined release artifacts that could easily be rolled back.

• Site reliability: eliminated the riskiest single points of failure and supported greater (and elastic) capacity.

• Observability: used DataDog to collect system, application, and business logs & metrics. This enabled the team to proactively & efficiently address issues.

Joined a small team to manage O(tens) of 100-node Elasticsearch clusters that housed O(terabyte) documents of O(petabyte) in total size.

• Used Salt to provision O(hundreds) of physical machines and worked with on-site operators to trouble-shoot hardware issues (eg failed disks).

• Used Kubernetes to provision Elasticsearch clusters & supporting systems such as Kafka.

• Developed custom sw to assist us with hardware failures & scale up & down events.

Added some SWE spice to a SYSADM soup:

• Developed a web-based tool to automate the provisioning of servers in environments lacking DHCP and PXE (no layer 2). By leveraging iPXE over HTTPS, the team was able to automate the OS install using Kickstart. Not only did this save time & frustration when working with clunky ILO & IPMI interfaces from across the globe, it eliminated problems with stale & inaccurate information in our asset management system.

• Architected, implemented, and operated a medium sized ELK stack for metrics & system logs. This allowed us to diagnose issues more efficiently, identify abuse more quickly, and be more proactive to capacity pressure.

• Developed a log shipper in node.js, which at the time was the only known such tool to support aggregation (roll-ups) at the edge. Without such functionality, trying to make sense of the multi-line system output proved too onerous.

• Operated on-premise GitLab instance. Introduced CI pipelines to the team, which improved the reliability & maintainability of projects using build pipelines with automatic QA gates and deliverable artifacts (RPM's, Docker images).

After the highs of a successful acquisition, I failed at my own venture but learned valuable lessons in entrepreneurship. I then began helping organizations with their transitions to AWS. I also helped automate, secure, and optimize their software delivery pipelines.

I graduated from ASU with a bachelor's degree in Computer Science. Held engineering roles across web, infrastructure, and startup domains, including:

• A NewTimes, I improved site reliability & performance by introducing caching, load balancing, and replication.

• At DataBill, I developed a high test-coverage framework for the primary business process which increased developer productivity and resulted in fewer errors at the end of the production pipeline.

• As the junior founder & principal engineer of Edgeos, I iteratively improved the product until an acquisition.