Designed, scoped, and developed a best-in-class product to replace emails, text messages, and PDFs with data-rich, mobile-friendly experiences for Commercial Real Estate brokers and their clients - Rebuilt existing application, overhauling the entire frontend codebase and replacing >20k lines of code with industry standard libraries and techniques - Expanded the application from original vision, adding functionality for custom PDF generation, website white-labeling, and a suite of portfolio management tools for brokers - Designed and implemented new deployment infrastructure with automated and horizontally scalable solutionsCo-developed business, product, and sales strategies with CEO - Led hiring process for engineering and early product hires and provided thought leadership across organizations - Key member of hiring process for sales/customer success teams - Advised on fundraising strategy and capital spend, contributed to meetings with investors and prospective clients

- Mentored junior members of the team- Implemented a test suite to aid in faster development and deployment of machine learning pipelines and jobs- Developed a front end application to provide insight to machine learning engineers about long running tasks- Maintained a Kubernetes-based system for deploying various applications including Jira, Gitlab, and the in-house machine learning platform- Managed the above Kubernetes-based system with Helm, Helmsman, and Cloudformation

- Identified pain points for Data Science and Application teams, planned and proposed solutions, gathered stakeholder feedback, and developed, deployed and maintained apps and infrastructure to address cross-functional needs- Devised, implemented, and maintained Drizly's event stream, built using (AWS) API gateway, Lambda, Kinesis (and Kinesis Firehose), S3, Redhsift, Spark- Leveraged Drizly's event stream to analyze checkout funnel and reveal key business insights to drive improved customer experience and conversion- Devised, implemented, and maintained Drizly's predictive model services infrastructure using Docker containers to spin up predictive models and expose them via a simple API to our internal network. The containers were orchestrated with ECS on Fargate.- Devised, implemented, and maintained a system to automate job processing via Cloudwatch alarms, Lambda and ECS on Fargate

- Led team meetings including sprint plannings, mid sprints, retros, and scoping sessions- Guided a team of 3-5 engineers on a product roadmap- Collaboratively constructed and carried out an engineering roadmap- Mentored more junior members of the team- Planned, scoped, and delivered features and fixes as a team

- Augmented and maintained a Rails-based web app hosted on AWS (EC2, S3, DynamoDB, SQS, Lambdas)- Created and maintained data migration/ingestion jobs (including some work with Spark)- Built out front end functionality using ReactJS- Created a processing pipeline to merge datasets using standard python and SKLearn for implementing and conditioning multiple classifiers - Facilitated the implementation of a predictive model service using the above processing pipeline as a platform- Implemented a click tracking pipeline using AWS' API Gateway, Lambda, Kinesis Stream and Firehose, S3, and Redshift

- Refactored Wayfair's front end browse code to remove legacy code while maintaining functionality. Organized using Tungsten, our javascript framework built on Backbone.js.- Owned various storefront changes utilizing Wayfair's in-house MVC framework in PHP, Javascript, Mustache templates, MSSQL, Vertica- Built out and maintained a testing framework (in PHP, relies on RabbitMQ for inter job communication), which allowed inspection of Wayfair's pre-search changes before deployment and integrated with an in-house data capture tool. This enabled us to capture the value of various pre-search layer changes.- Responsible for a "pre-search" application layer that enabled us to build more informed SOLR queries. Built in Python.- Helped troubleshoot bugs and add new features to a distributed email queuing system based on RabbitMQ with Apache Storm- Augmented and maintained the full stack of a graphical interface for displaying timseries data using Elasticsearch and MySQL as well as PHP and Javascript/JQuery.

- Continued work on a graphical interface to visualize data from multiple sources in one place- Modified and maintained a script to sync over one million metrics to elasticsearch indices- Maintained a strong work rate while remote from the team and finishing my undergraduate degree

- Worked on various projects from UI to security across the full web stack

- Worked with a team of undergraduate and graduate students to develop a Graphical User Interface Debugger called GUITAR- Developed a visualizer using PyGTK (version 1.0) and Java SWING (version 2.0) by assessing the needs of both the team and the end user