I've worn a lot of hats during my time at Sift -- overall I have spent most of my time working on distributed Java services for our post-ML products, both creating new, and scaling the old.

Some notable projects/accomplishments include:

• Voted "Get Sift Done" award by my peers for acting as a cultural role model

• Developed several iterations of a key new customer reporting product for aggregate insights into customer data, from database design to API. First backed by partitioned PostgreSQL, and then migrated to an AWS Kinesis + Firehose + Redshift pipeline.

• Scaling up and sharding of virtual queue infrastructure for Sift's Workflows rules-engine backend.

• Performance analysis and tuning of legacy (1.7) ElasticSearch infrastructure, followed by a live migration from 1.7 to 7.0

• Designed and implemented efforts towards workload isolation by splitting out several monolith services into component services.

• Relevant software: Kafka, HBase/BigTable, AWS/GCP, Zookeeper, Memcached, PostgreSQL

Machine Learning > Fraud

Full stack development of scalable mobile application backends from inception to launch. Notable projects include:

http://channel.wtf

Designed and automated closed captioning of arbitrary mp4 inputs via AWS Lambda pipelines and IBM's Watson API

http://joinchit.chat

Full migration of application and network architecture of realtime chat app from NodeJS to Java backend

Relevant technologies:

Java/Maven, AWS (RDS, SQS, SNS, Elasticache, Lambda + various networking/VPC stack), Google Cloud, Google Guice, Redis, MySQL, DropWizard, Terraform, Ansible, Vagrant, NewRelic, git