May 2015 – Jan 2022 · 6 yrs 8 mos · London Area, United Kingdom

• I had the opportunity to work for Twitter in two regions (London & San Francisco) and two major groups (Revenue & Consumer). During that time I helped drive major projects within the company, started multiple new teams to tackle previously unrecognized challenges, and managed orgs with 20+ people on a variety of projects, including:
• Kickstarting a rebuild of Twitter's runtime search infrastructure, beginning with user search and expanding to Tweet search
• Launching Twitter's live events product which required building scalable, resilient infrastructure that could support video streaming and tweet distribution with very spiky traffic and across a variety of languages and regions.
• Twitter's ML Feature Registry was born out of a prototype built by my team in conjunction with the Revenue Data Science team. That initial project became the focus of an entire team and turned into a key piece of ML infrastructure at Twitter.
• Overhauling how user data was stored and protected at Twitter. The user visible portion of this was Twitter's first privacy dashboard, allowing users to see what data Twitter stored about them, but under the hood this involved working closely with legal and privacy teams at Twitter to ensure the data we stored was kept in safe and compliant manner, which later benefited GDPR efforts at Twitter.
• Building out Twitter compatible infrastructure in AWS & GCP. Our goal was to make Amazon and Google clouds operate as seamlessly as possible for Twitter engineers. This involved not just technical challenges but also work to ensure we met financial and security compliance standards.
• I was also involved in other projects, including:
• Rebuilding career ladders across the technical disciplines to ensure consistency and simplicity.
• Co-leading the London Engineering team's Recruiting & Representation function.
• Designing rubrics and questions for Twitter's software engineering interview process, to provide a common approach across all teams.

Dec 2013 – May 2015 · 1 yr 5 mos · Burlingame, CA

• I worked with a great group of engineers building out the next generation of TellApart services, including:
• Core services built on top of the lambda architecture (hadoop, kafka, s3, voldemort and spark)
• Manageability toolsets for a service oriented architecture (service discovery, monitoring, configuration management)
• Reporting and data pipline (hadoop, spark, shark, kafka, hive and cascading all play a part)
• TellApart was a great place to work with a top notch engineering team! We believe in moving fast and taking a data driven approach to the work we do. The data is big (over a petabyte in storage) and the scale is massive: we build services capable of scaling to thousands of requests per second with latency below 40 milliseconds.
• We've been acquired by Twitter!

Apr 2011 – Nov 2013 · 2 yrs 7 mos · Sunnyvale, CA

• I was recruited to proofpoint to help build a next generation anti-spam, anti-malware, anti-phishing product. My contributions ranged from individual contributor and tech lead at the beginning to managing and leading three distinct engineering groups, including the core anti-spam group around which proofpoint's business was built. During my tenure there, I lead and built a team of engineers that launched that next generation (and very successful) product. To do so, we had to:
• Build a data collection pipeline capturing metadata from hundreds of millions of email messages. The raw data volumes under storage exceeded a petabyte by the time I left.
• Create and automate analysis jobs to processed the message metadata to identify spammers, phishers and hackers
• Provide real time dashboards and alerts to customers about the threats swimming in their email streams, including specific information about who was targeted and with what type of attack
• Invent a real time scoring service capable of scoring thousands of messages per second
• To do all the above, we used mainly java based webservices built around proofpoint's open source Platform libraries, along with healthy doses of hadoop, hive, redis and cassandra. For scoring and evaluating emails we used machine learning techniques including neural nets and logistic regression.

Jan 2010 – Nov 2011 · 1 yr 10 mos · Palo Alto, California, United States

• At Occam we leveraged open source search technologies, coupled with proprietary AI to build a legal search engine. I was responsible for building the product, and building and leading our dev teams.
• Funded by Sequoia, Silicon Valley Angels

Aug 2008 – Jan 2010 · 1 yr 5 mos

Jan 2005 – Jul 2008 · 3 yrs 6 mos

Jan 2001 – Jan 2002 · 1 yr

Jan 2000 – Jan 2001 · 1 yr

Jan 1999 – Jan 2000 · 1 yr