Managed a team of 6 SREs focused on physical datacenter provisioning, puppet, kubernetes, incident response and observability and monitoring.

Responsible for quality and performance of the software and systems architecture for Postmates. Identify scaling bottlenecks. Build solutions to keep up with growth of our systems. Work with other engineers to build better software.* Designed major shift in site infrastructure to use Kinesis and Redshift to move us from a single monolithic Django app, into a services-based architecture.* Oversaw major initiative to address monitoring and metrics deficiencies in site architecture.* Oversaw migration from salt-based infrastructure and deployment to an AMI-based immutable infrastructure using Consul.

Set technical direction for major projects, with a minor focus in site scalability. * Designed, documented and help build public API to enable 3rd party developers to use our awesome delivery platform. * Split-up large Django project into multiple "sites" to allow for varying configurations. * Built out site-infrastructure to allow us to scale database usage across multiple replicas. Used postgresql, pgpool2, pgbouncer and lots of AWS.* Instituted several software development processes such as formal documentation for projects and design proposals.* Major focus on hiring and interviewing.

* Scaled out a small Django app running on a single instance to a fully automated infrastructure in AWS. Ubuntu/PostgreSQL/Memcache/RabbitMQ/Celery* Integrated several of my own open source projects for deployment (FTL) and logging (BlueOx)* Built a SQL-based metrics system for monitoring key performance indicators.* Built product search system using AWS CloudSearch* Built javascript front-end for product search, including auto-complete.

Developed a variety of products starting with personal data space (Retrosift.com), a group photo album product (Lensmob.com) and more.Developed our backend infrastructure, from deployment to scalability with AWS. Started several open source projects during this time: Dynochemy (Python library for using DynamoDB), BlueOx (a python logging aggregation system) and FTL (a deployment system)

Managed the Revenue Team (Ads and Billing) growing it to a team of 8. Later, switched to Infrastructure team where I managed a team of 5+ (depending on interns) focusing on site scalability and reliability. Still wrote plenty of code.

Wrote lots of Python code for a variety of projects. * I developed a Billing and Payment Processing system that replaced a series excel spreadsheets into a full fledged scalable system that processed millions of dollars a month. * Additionally I developed and scaled the ad serving and targeting system. * Wrote a bulk email system (the weekly yelp) that delivered millions of emails each weak, complete with click tracking and reports.

Worked on the Serviceguard product, a high-availability clustering product developed in the HP-UX lab.Considered a mission critical application by many high profile customers, Serviceguard is a complex, near realtime distributed application that required a level of quality and technical expertise. I learned everything I know about real software engineering from this job.My focus with this product was on networking and security. As security specialist, I did security audits of existing features, planned new security functionality, wrote documentation, participated in security patch releases and gave several instructional classes to other developers about security and Serviceguard.In addition, I was the lead developer for our internal web applications which automated our release and quality processes. This was my first foray into python, developing a full web framework before Django or related existed.

Combination of systems administrator, data scientist, web designer and programmer.My primary responsibility was to manage web traffic analysis for MGM.com properties. This meant maintaining redundant WebTrends machines (Solaris and Linux), managing the work flow of retrieving logs via FTP and loading them into WebTrends. Recovering data when these fragile systems failed. I was also the goto web traffic analysis expert so I'd setup reports for various questions the marketing people at MGM had.Eventually I wrote a fairly extensive python application for doing log management. It automated the downloading, recovering, pre-sorting and data cleanup that was necessary to make my job less painful.

College job. Started off as a web designer / HTML monkey for a variety of their websites. Eventually convinced my boss to let me try my hand at Perl coding by adding a feedback form. By the time I left I built 'stocksport.com', a educational stock market trading platform, all in Perl.