Day AI is the Customer Intelligence solution that puts customers at the center of every decision by reimagining CRM for the AI Age.

We combine the features of a Meeting Assistant, CRM, and Knowledge Base into a single unified solution that replaces the manual, complicated interfaces of legacy CRM, and keeps you connected to what customers want.

After my previous team launched a re-imagined CI/CD experience within Netflix and drove it to product-market fit, I worked with other engineering leaders to re-organize the Delivery group around scaling adoption of the new experience and expanding it to more of Netflix's engineering ecosystem.

As a result, I took over the newly formed End Developer Delivery team which consolidated ownership of all the major developer-facing experiences for CI/CD and aimed to unify them into a single, seamless workflow.

I led the Managed Delivery team, which aimed to rethink how Netflix interacted with continuous delivery. The team operated as a startup within the existing Netflix Platform organization, experimenting with a new CI/CD experience in close partnership with representative customers across the major engineering functions. We also regularly met with industry peers to drive a common language and set of concepts for abstracted or "managed" CI/CD at FAANG scale.

After some research and development, we launched in closed beta and eventually found product-market fit within several large engineering departments. By the time the team was spun down to move into a more scale-up focused configuration, the product was the default for new services at Netflix and was growing at over 100% YoY.

I worked on Spinnaker, Netflix's deployment, cloud infrastructure management, and pipelining/orchestration platform used by engineers across the company.

My feature development focus was on the UI, but I was also on the PagerDuty rotation for the overall system including a constellation of Java services that ranged from a Netflix-scale cloud infrastructure cache to the core orchestrator that ran workflow steps in a JSON-based graph/DAG specification. Spinnaker was open sourced in 2015 so you can read all about the services involved in its operation on the docs site (link below). That rotation also included Slack-based user support for all the engineers at Netflix who needed help, ranging from setting up deployments correctly to helping remediate critical streaming outages.

The time I spent in this role was incredibly challenging, but was also one of the most illuminating and educational experiences I've had in my career. Staffing user support and deep diving with engineering teams across the company was effectively a way to do a tour of duty with every product function at Netflix, and that combined with holding the pager for Spinnaker gave me a 2 year crash course in what it looks like to build distributed systems at Netflix scale.

While working on email client extensions at HubSpot, I came upon the work Streak was doing with a developer-facing project called InboxSDK. They had accomplished an incredible amount of deep work to manage UI injection into Gmail, including robustness to Gmail UI deployments that could break extension code, and automatic reactive handling of UI transitions/interactions. Most impressively, they'd made a lot of progress on the problem of coordinating UI injection across the many different email extensions a user might have installed — effectively a distributed consensus problem — which had been one of the most frustrating parts of email extension development for years.

We quickly transitioned much of HubSpot's bespoke extension code to InboxSDK, and as I got to know the Streak team better I decided it would be an exciting adventure to move to San Francisco and go even deeper on some of the tough platform engineering work I'd already gotten a taste of. What ensued was an incredibly learning-packed year, I probably went deeper on the browser platform/runtime during this time in my career than I'll ever go again. The SDK also supported millions of email clients through companies like Dropbox and Giphy using it for their extensions, which helped prepare me for the scale I would soon be working at in the next chapter of my work.