Taking care of ill family member

Reduced CarGurus website JVM memory usage by ~2GB (~12%) by implementing size-bounded LRU lazy-loaded cache + Memcached, also reducing database load on production site deployment and improving app startup time by up to ~3 minutes (20% improvement)Saved ~1TB of memory in our dev/test/CI environments by adjusting JVM values for deployments of multiple services, removing a resource constraint on setting up new hire devboxesDrove, implemented, and deployed time-critical security updates for services across the engineering orgSME for the CarGurus site build and deployment pipeline, responsible for training new eng hires on deployment processes and health verification across various metricsExpert advisor and escalation on-call for rotating team of 10 volunteer site deployment monitors

Primary SRE for entire Consumer Engineering department (~40 ->60 engineers)Tuned JVM memory allocations and garbage collection settings for primary CarGurus site and backing services to eliminate stop-the-world garbage collection pauses leading to cascading outagesRan experimentation, testing, implementation of switch from CMS to G1GC for primary CarGurus site + backing servicesMaintained and updated deployment pipeline for primary CarGurus site + backing services (Rundeck + Python + basic bash scripting)Trained and guided engineers across the company on production hotfix processesLed troubleshooting, communication and remediation of numerous production incidents

Set up alerting and monitoring for the CarGurus website and backing services using Prometheus/AlertManager/OpsGenie/Grafana stack (Terraform + Chef + Ansible + AWS) in preparation for standing up an SRE team.Fixed and hardened CarGurus site production deployment scripts (Python + Rundeck)

Dual role, 50/50 time splitReliability Specialist (Site Reliability Engineer by another name):- Triage, perform root cause analysis, and solve specific classes of production problems affecting our core website (Generally when either the cause is not immediately obvious, or ownership of the offending code is nebulous)- Proactively detect latent production issues that may be flying under the radar, make them visible, and drive the solution- Drive good development decision making across multiple dev teams with a focus on stability and scalabilityCustomer Experience, Studio- Create dashboards and thresholds in multiple analytics apps (Sumologic, New Relic, Rollbar) to appropriately alert studio team to production issues as soon as possible.- Ensure Studio-specific architecture is performant and able to handle the load for our peak traffic holiday months- Maintain C# back end architecture for our customer facing product design JS app- Updated our C# content rendering pipeline to enable launching several new product types that rely on new printing technologyPreviously:VBU Product Design Technology- Developed a .Net Core microservice connected to an ElasticSearch cluster to serve design data for user facing new feature testing in production.- Working to modernize legacy code into RESTful microservices on AWS- Maintain our C# content creation pipeline- Maintain our Adobe InDesign plugin suite

DCS New Development team- Backend developer for new internal client contract setup web app using DB2/Hibernate/Java/Spring/Ext JS technology stack interfacing with legacy database- Designed and developed Java RESTful backend to handle updating zip code zones with over 20k entries, turning a weeks long manual spreadsheet process into a 5 second CSV upload

DCS Startup team- Developed custom C# integration solutions to process order data feeds from new customers with a variety of formats (XML, JSON, CSV, Excel)- Created an auto refreshing grid overlay of truck data in our Trucks on a Map Java web app using the Dojo JavaScript framework- Primary point of technical support for client worth $40 million in annual revenue and 3,000 orders daily- Implemented C# integration solution for manufacturer’s XML data feed with less than 2 weeks’ notice to support new customer’s Black Friday initiative- Subject matter expert in our order and routing systems (multiple C# windows services + C# WCF web services with IBM MQ messaging)

DCS Rapid Response team- Provided 24/7 on call support for applications and processes totaling over 2 million lines of code in Java, C#, COBOL, VBA .Net, JavaScript, and DB2 SQL- Implemented defect fixes, small features and SSRS reports for software projects scoped under 120 hours in C#, Java, JavaScript, and DB2 SQL- Recovered over $430,000 in missing billable revenue by tracking down and fixing defect in the Drools rule engine of our Java invoicing system, then saved over 80 hours in additional invoicing work to facilitate prompt billing of customer