* Moved candidate data pipeline from a cluster of loosely-affiliated REST Applications to a modern data pipeline based on Spark and Airflow* Married multiple data sources to produce rich, up-to-date candidate profiles with great DEI signals* Implemented the company's first Terraform installation and evangelized configuration-based infrastructure to the team* Helped implement Scala best practices for both the Spark pipeline and Play Framework-based applications* Analyzed & managed infrastructure spending on EMR, S3 and other AWS products

* Served terms as lead of Slack's #Abilities ERG and #Out ERG; organized charity events, programming and lectures for staff, mediated disputes, and helped constituents to feel happy and comfortable at work.* Led a team in launching Slack's Data Discovery Service, which ingests SQL and deduces data lineage information based on an ANTLR SQL grammar parser.* Designed high-performance spark applications, including a solution which uses Spark's GraphX library to flatten the aforementioned lineage information into a flat dataset.* Designed airflow operators for custom functionality such as data expiration.* Contributed to core logging infrastructure code in Slack's webapp.* Helped educate coworkers about application design principles, such as the 12 factor app and structured logging.* Migrated, split, combined and retired services.* Lead a team in migrating the internal data portal to an OpenSearch (Amazon's version of ElasticSearch)-based solution.* Wrote terraform and chef code defining resources for the Data Tools and Data Ingestion teams.* Worked with our Cloud Infra team to write new features for Slack's go-based kubernetes management system, Bedrock.

Migrated analytics ETL to Spark. Cut costs on AWS infrastructure by reducing redundant infrastructure, eliminating wasteful over-usage of s3.

Rebuilt backend team from scratch. After being put in the position of being the only person who could address certain problems, I've built a small, passionate, highly skilled team that can support production systems with little direct intervention from me on a day-to-day basis.Trained developers in use of Jenkins, Docker, Mesos, Marathon and AWS: how to launch and configure applications, how to build containers, how to cobble together a CI system, and how to troubleshoot them in production. Designed many integrations with third-party systems, including DMPs and content feeds.Lead migration to managed AWS services on multiple fronts - Cassandra, HBase and Redis to DynamoDB, Ambari Hadoop to EMR - realizing significant cost savings and simplifying operations.Architected an Amazon Aurora-based service which provides both fast and up-to-date analytics queries for a variety of objects. Rearchitected a Redshift-based analytics cluster for much better performance, and integrated it with Looker, a data visualization PaaS.Implemented metrics and monitoring systems: built out metrics infrastructure on a Graphite/Grafana/StatsD stack. Wired StatsD into DataDog, a subscription metrics PaaS with the ability to trigger alarms on PagerDuty. Implemented PagerDuty alerts in the case of fatal errors where practical. These systems, in concert with sensible operational polciies, have lead to fewer and less severe interruptions in service across the board.Supported an enterprise-class, high-performance ad search engine based on Lucene: wrote documentation and user guides, reviewed code, fixed bugs and built supporting systems and instrumentation.

* Built and launched an ElasticSearch-based ad server (CORE), capable of dynamically targeting audiences to content using simple search queries at scale, as well as a catalog-based ETL pipeline and monitoring infrastructure to ensure reliability.* Piloted methods for using TDD and CI in a microservice ecosystem.* Developed policies and guidelines for effective development in a Java/Spring Boot-based microservice ecosystem.* Primary Node.js developer: I have expanded the capabilities of a Node dynamic ad server (Weather forecasts, behavioral targeting, geolookup) and introduced several performance optimizations, including integrating it with a fast UDP geo server and implementing multiprocess clustering. Currently researching porting server to Go or Java.* Rails: I ported many existing database queries to ElasticSearch, and in the process learned more than I ever wanted to about the ES query api. Wrote test helper to generate ES records using FactoryGirl-like syntax. Used lazy loading and Angular to make app more responsive. Optimized many MySQL queries.

* APIs and applications for set-top streaming boxes (Roku, Boxee)* Flash Video player and ad serving infrastructure* PHP front-end code

* Website and web API (Java Servlets)* Data processing and analysis (C++, XSLT, Bash)* Project management* Technical writing and communications