Senior Software Engineer on the Business Insights team at Stripe

Lead developer of GIPHY’s centralized moderation pipeline, which replaced several existing businessprocesses and streamlined all content moderation decisions. Backed by Celery (a python task queue), this pipeline is highly customizable, scalable, and supports automated and human-in-the-loop moderation. Trained several developers to contribute to the pipeline. Processes ∼500,000 moderation related tasks per day.Designed and developed the Hash Slinging Slasher, a hash matching pipeline which tracks files thathave been deemed unsuitable for GIPHY and prevents those files from being reuploaded, catching and blocking about 30 uploads per day. This pipeline uses both MD5 hashes for exact file matching, and perceptual hashing such as PhotoDNA and PDQ for matching modified files. Currently integrating with Meta ThreatExchange for sharing signals across organizations, andthe Global Internet Forum to Counter Terrorism hash sharing database to greatly improve Trust and Safety.Integrated with NCMEC PhotoDNA perceptual hash database to enable automated reporting ofknown illegal content to law enforcement without requiring human moderator intervention. Reports about 30 pieces of content to law enforcement per day.Built EMR jobs for aggregating and processing a wide variety of Trust and Safety data.Tech Lead for multiple feature development projects to support content creation, ingestion, editing, and moderation. Established teams for building out features, organized sprints, project planning, and architecture design. Mentored junior developers as part of design and implementation process.Maintain and improve core backend video renditions processing. Designed, developed, and currently maintain a secondary renditions pipeline for rendering non-essential renditions offline using a queue-based (SQS) approach.Contributed bug fixes to FFmpeg - https://engineering.giphy.com/fixing-bugs-in-ffmpeg-gif-encoding/

As a member of the Data Engineering team, I maintain and improve several applications involved in the movement and processing of high-volume data. I participate in the on-call rotation, and have diagnosed and fixed several major outages of our platform. As part of a major project, I designed and implemented exponential backoff and proportional control algorithm for Kafka forwarder application, to reduce latency and improve network congestion.

Redesigned message parser generator, a major component of the build process for most AppNexus applications, to make it modular, readable, and testable. Experimented with both software and hardware prefetching to improve speed of the bidder application.

Built automation software on top of Tomcat and Maven, designed to run on BNY Mellon proprietary cloud platform; increased ease of access to open source software. Discovered and reported security hole in email system.

Designing and developing iOS app that provides assistance to potential customers and legal information to attorneys working at the firm