Platform Engineering at Dandy

Worked to improve Wayfair engineering efficiency by developing systems for creating automated pull requests, templating new applications quickly, and cataloging existing software.

Worked collaboratively to design, build, deploy and maintain a change-propagation system enabling any engineer to easily create PRs en masse across our decoupled ecosystem, leveraging FastAPI, Tekton, a relational DB, Kafka and React (highlighted in episode 357 off the Python podcast podcast.__init__)

Deployed and scaled Renovate (an open-source tool for automating PRs for updating versioned software dependencies) to 1k+ repositories in Wayfair’s ecosystem, making and facilitating open-source contributions when necessary

Maintained tooling for templating new applications, allowing engineers to create, deploy, and decommission applications in less than 5 minutes

Refactored Wayfair’s Product Search data pipeline from SQL-based batch system to multi-source event-based low latency data update system supporting 500+ search indexes, leveraging NoSQL datastore Aerospike, Apache Kafka, Dropwizard and Google Guice

Supported 1000+ instances of Apache Solr across multiple datacenters on-premise and in Google Cloud, utilizing the Elastic stack, Datadog, InfluxDB and Jenkins

Optimized Solr instances for query performance and data update speed, leveraging caching strategies, cluster rebalancing, commit strategies and index merging policies

Designed, built and deployed Wayfair’s keyword search spellchecker, written in Python based on Wilbur Et. Al.’s spellchecker at Pub Med

Owned and rebuilt Search Tech’s monitoring and alerting, reducing engineering support time and increasing system resiliency

Taught students valuable C++ and Python programming concepts

Graded 7-10 assignments submitted by students weekly