Mar 2023 – Sep 2025 · 2 yrs 6 mos

• 2024 - 2025 Director of Data Platform leading 100+ organization; Data Infrastructure, ML Feature Platform, Training Data, Experimentation Platform, Analytics Platform, Data Engineering, Data Governance, Generative AI tooling, and Data Surfaces. Platform powers data needs for all Snap Engineering, Applied ML, Product, Growth, Data Science, and Sales teams. Scale: exabytes under storage with 1 exabyte ingested weekly. 9 figure infrastructure budget. One of the largest experimentation platforms in the industry with 10K+ active experiments at any time. Operating 8 legally required compliance programs with >700 ITGC controls, and responsibility to protect 10K/10Q metrics.
• 2023 - Led DevProd as Sr. EM. Directed 23 engineers across CI platforms, source control, builder provisioning, and mobile app CI. Delivered Copilot to all of Snap, and built agent framework.

Mar 2019 – Mar 2023 · 4 yrs

• 2021 - Nov 2022. Expanded role of Data Infra to manage ML, operational, and analytics with the mission to become the singular data platform for all of Snap. Broke apart and rebuilt existing monolith infrastructure into 3 core platforms: collection platform, processing platform, and datalake. Grew team to 35+ engineers to support additional work loads and internal teams. Leading multiple managers and Staff/Principal level engineers. Key Technologies: Spark, Kafka, Iceberg, Airflow, Atlas, BigQuery, PubSub, and Kubernetes.
• 2020 - 2021. Reorganized data teams around company under a Data Infra org with 20+ engineers. Primary focus on business analytics. Leading the overall org and redefined how data teams are setup; shifted 2 large backend teams owning everything into 4 distinct vertical teams from data collection to data-at-rest.
• 2019 - 2020. Rebuilt logging SDKs team and lead re-arch of Snap's entire mobile logging infrastructure with a team of 7. Fixed the most critical piece in Snap's data ecosystem at the time. New mobile logging infra scaled with company as a result - particularly on Android with rapid adoption of Snapchat in emerging markets with low-connectivity devices.

Oct 2022 – Mar 2023 · 5 mos

• Built out a passion project in Consumer EdTech I’ve had for a while. Teaching kids real Java coding through simple daily practice. No mindless games, no abstraction layers from real code, no IDE setup needed. Have your kid write real working code within 10 minutes.
• We offer a a full fledged web-based IDE augmented with thousands of exercises that are designed to keep coding skills sharp for kids aged 8-18 from all experience levels. 30 mins a day is all you need.

Mar 2018 – Mar 2019 · 1 yr

• Led store health and device data streaming platform for Amazon Go. The platform manages petabytes of streamed data (video, metrics, load cells etc.) across thousands of devices while providing value-add services such as: monitoring, access control, anomaly detection, and retention.
• 3 patents filed with USPTO related to sunlight detection system and image quality monitoring.
• Led teams to deliver device monitoring and camera infrastructure to power the seamless launch of 9 Amazon Go stores (4 Seattle, 3 Chicago, 2 San Francisco) in 2018/19.
• Expanded and drove roadmap to drive down OpEx cost for infrastructure through setting clear goals and delivering: data compression on device streams, serverless architecture, and optimizing on data store indexing and usage patterns.
• Expanded initial team of 4 SDEs in Q3 2017 to 3 teams across 22 engineers by Q4 2018.

Dec 2015 – Mar 2018 · 2 yrs 3 mos

• Founding member of Store Health team that designed and implemented device signal monitoring platform to constantly measure the overall health state of an Amazon Go store in real-time. Solving challenging health problems such as sunlight detection and characterization, anomaly detection on data streams, and visualization of hardware component failures.

Apr 2015 – Dec 2015 · 8 mos

• Working on Amazon Video Precognition Team
• Delivered 'Watch Next' carousel which is the top performing carousel for stream attributions across all devices. Involved in all aspects of delivery from product requirements gathering, project management, technical design, and final implementation.
• Developing Machine Learning platform to generate smart video lists, and next title recommendations for Amazon customers.

Jul 2013 – Apr 2015 · 1 yr 9 mos

• Working on Amazon Video Catalog & Playback teams
• Delivered 2014 Ultra HD launch for AIV in US/EU. Designing and implementing much of the backend solution to support Ultra HD offers, ownership, H265 asset delivery and content protection.
• Worked on the full stack for AIV platform services, from catalog publishing, playback/DRM, and video purchasing/ownership.
• Contributed to Amazon Instant Video in Japan in 2013, UK & Germany in 2014; building out an entirely new catalog publishing stack to support launch.

Mar 2012 – Jun 2013 · 1 yr 3 mos

• Working on Amazon Video Support Engineering Team
• Developing operational, alarming, reporting and troubleshooting tools for Purchasing, Ownership, Discovery and Catalog systems.
• Worked in London post LOVEFiLM acquisition and migrated LOVEFiLM stack to Amazon. Transferring AIV knowledge and web services to London-based teams.
• Identifying patterns in customer contacts and track down customer impacting issues.

Jun 2010 – Mar 2012 · 1 yr 9 mos

• Built & support voter identification application that interfaces with call centre used by 100 agents nightly.
• Developed www.panelcloud.com, a fully scalable online panel that has a full rewards engine.
• Worked closely with larger clients to develop long-term, efficient data-management processes. This included transforming data returned from our call centre into acceptable format for upload into their existing databases.
• Moved entire company’s IT services (Call centre, website, call recordings server) to Amazon Web Services.

May 2009 – Jun 2010 · 1 yr 1 mo

• Improving upon existing software for www.whistler.com written in C# and .NET 2.0. Primarily, improving hotel booking engine speed and accuracy.
• Built and deployed touch-interface factory floor quality control software for Axion Power.
• Headed the media team for www.kidobi.com, a children’s media website, building an Adobe Flex Player that communicates back and forth with a .NET/C# media engine as well as integrated video streaming to Amazon CloudFront services. The video player was augmented with parental controls, back-end video analytics and complex playback options.

May 2007 – Sep 2008 · 1 yr 4 mos · Toronto

• Web based Production Scheduler/Optimization software written in C#; used primarily in refineries but can be used to model and optimize any production facility.
• Developer for user interface and presentation layer. Coordinated extensively with Toronto team and test team in India. Pushed initiatives and technologies such as AJAX to be implemented for better user experience.
• Writing and improving existing SQL stored procedures. Writing SQL to generate views that are used to improve query performance.
• Implemented user-centric features such as: load on demand grid controls, caching user preferences, and improving UI layout to be more intuitive.