Facebook Reels: feed delivery infrastructure

• Optimizing for Reels feed delivery latency and reliability

Facebook Live Video: product infrastructure

• Responsible for foundational livestreaming functionality: data model; entity state/lifecycle management; privacy & permissioning; building internal APIs for product teams to build on top of; interfacing with lower-level livestreaming infra; maintaining data integrity & repairing data corruption; APIs for 3rd-party livestreaming apps

• Improve system design and architecture to increase reliability and efficiency of our product. Enable other engineers to move fast safely by designing code that funnels them towards safer patterns and practices, and providing convenient, accurate-to-production test utils.

• Technical leadership of my direct team by formulating project plans to address problems we encounter, and making design decisions & tradeoffs in our projects.

• Provide technical guidance and expertise to other teams across the videos org & across the company on topics such as framework improvements; pre-empting impending disasters; tech talks; Q&A + system design feedback to engineers on their projects

• Optimized for latency vs capacity concerns for sending high-volume notifications in real-time, in the presence of thundering-herd behavior by broadcasters.

• Operational management and continuous improvement of live video service & reliability for our users through oncall rotations

Facebook Live: product

• Managed the health of live video notification ecosystem, one of the most important distribution channels for live video content in Facebook. Managed competing priorities of stakeholders who want to send more notifications to produce more watch time against priorities of stakeholders who seek to preserve health of user experience as measured by clickthrough rate and notification usefulness survey.

• Development of multiple core services (written in Java/Mongo, Go/Postgres) in a microservices architecture

• Worked on: service resiliency initiatives; breaking synchronous processing out to asynchronous to speed up request latency; decoupling services from each other; eventual consistency among services that have separate data stores; application monitoring / logging / distributed tracing

• Wrote and improved scripts and libraries used by devs across the company.

Keywords: Java, Go, Python, Thrift, gRPC, AWS, MongoDB, PostgreSQL, Serverless, Lambda, Git, Gerrit, PubSub, Monitoring and logging, Datadog APM integrations, Sentry, Asynchronous processors, DevOps, Security

As engineering hire #2, I built various components of the Swipecast app's backend: server endpoints, codifying of business logic, login & security, payments, database management, server scalability, admin dashboards, automation, and high-level system design decisions.

Keywords: Ruby on Rails, Heroku, SQL, React.js, React Native, Redux.js, ES6, GitHub, Google App Engine, PHP

API integrations: Stripe, Pusher, Instagram, Facebook, Postmark, and more.

Proprietary trading in electricity derivative markets (futures, financial transmission rights)

Macro hedge fund trading rates, FX, and equities