🚨 Maintains critical site reliability platforms

Currently develops and maintains LinkedIn’s internal alerting, notification, oncall management, and automated remediation platforms.

💰 Cuts costs

Independently (as a side-project) scoped and executed infrastructure changes saving over $1.2 million per year in hardware costs by reducing the resource use of two large-scale applications, done in the time span of 2 months.

🧑‍🔬 Helps test high-stakes changes

Built request replay tooling to evaluate the performance and correctness of LinkedIn’s new metrics platform (leveraging distributed traces), allowing the team to debug and fix underlying issues before the migration, as well as measure progress to platform parity. Included building a pretty rad + intuitive Grafana dashboard.

📊 Leverages open-source technology

Created an internal fork of opentelemetry-collector-contrib to support ingesting distributed trace data (in a proprietary LinkedIn schema), allowing the use existing open-source technology built on OpenTelemetry.

🚀 Handles massive scale

Supported backend applications servicing more than 4 million requests per second.

💯 Maintains high availability

Maintained 99.9% availability of all online applications comprising LinkedIn’s targeting, ramping, and experimentation platform (T-REX).

🤖 Improves site reliability

Built automation to incrementally deploy experiments while monitoring and reacting to their impact on application health signals, reducing likelihood of incidents from insufficiently validated A/B test treatments.

📋 Builds alerts & dashboards-as-code plugin

Created (as a side-project) a GitHub Action to allow teams to manage their dashboards and alerts as code, facilitating better alerting practices through code-review as well as providing discoverable examples to be leveraged by others, improving oncall quality-of-life.

🔍 Prevents widespread outages

Evaluated production readiness of a service slated to be used by almost our entire fleet. Developed and executed load tests to proactively identify severe scale limitations, preventing the issue from occurring in production.

💰 Lowers user-acquisition costs

Created a browser registration funnel for our previously mobile-only application, reducing cost of user acquisition by 50% and paving the way for future investment into a cross-platform target for application development.

🔧 Automates the hard stuff

Created a Python tool to automate safe deployments of our legacy C# applications (including our main monolith), eliminating all deployment-related site-reliability issues. Included writing thorough documentation and training 30+ developers on tools use, which allowed engineers to confidently and safely deploy new code without fear of user impact.

🤝🏻 Builds appreciation culture

Built an internal Slack application for co-workers to show each other appreciation through coin emojis, dubbed ‘Earnin Coin’, which went viral internally. Regularly reached 100s of weekly transactions. Included a real economy, a weekly leaderboard to celebrate high coin earners, and was consistently touted as making employees feel better about giving and receiving help.

📲 Builds job-dispatching application

Worked as an independent consultant, building an SMS-based job-dispatching application for an industrial and oil-field safety services company using Node.js and Twilio. allowed the company to automate notifying (and provisioning) qualified employees for available jobs.

💬 Creates secure chat application

Worked as a fullstack developer to create a chat application for health-care professionals, facilitating the secure sharing of patient information between clinics and physicians within Canada. Built using Angular and Node.js, leveraging WebSockets for real-time communication and notifications.