A conversation
about ownership.

The project I'm most proud of is Driver AI at Lyft. I led the Android client from the initial specification through public launch. I'm proud of it because ownership meant bringing the engineering and operational teams together around the complete driver experience.

The product combined chat and voice, with dependencies on shared message contracts, backend behavior, and iOS. I wrote the Android specification, worked with the contract owner on additive chat and tile models, and implemented speech input, plan cards, and driving-aware controls.

I also brought client, backend, and data engineering together to align on what data we were gathering. Alongside the technical work, I coordinated with support and the driver team about the changes. That extended my responsibility to making sure the teams around the experience understood what we were shipping.

When duplicated messages and backend ownership gaps became launch blockers, I helped resolve them across teams. My direct responsibility was Android delivery, while the other owners retained responsibility for their systems. I stayed involved in the integration until the complete experience was ready to launch.

The product launched publicly, and the chat patterns were later reused. What makes me proud is connecting the implementation, data collection, and team alignment through to delivery. It reinforced how I approach complex projects: make contracts and responsibilities explicit, bring the right people into the decisions, and follow the work through launch. I can go deeper on the shared contracts, data alignment, or launch blockers.

There's actually a funny coincidence behind my interest. I recently moved, and my bed is by a window, so that side of the room gets colder. I already pay attention to my sleep routine—I take magnesium and L-theanine before bed—but room temperature is harder to manage. My partner has also mentioned that I snore. Eight Sleep was already on my mind when Gary reached out about this opportunity.

More broadly, I've experimented with supplements and different diets, and I've learned to value what is sustainable and works for me. Sleep is a consistent priority. I'd like to bring my engineering work into health technology because it's an area I care about personally.

The temperature control and personalization are what make the Pod interesting to me. It's also a significant investment, so I think about what would make it worth incorporating into my daily life: a reliable experience and value I can understand over time.

As an Android engineer, that connection between the app and a physical experience is compelling. My work on Android Auto reliability and shared mobile capabilities gives me relevant experience to bring, while learning this product's device and sleep domain. That's why this feels like a meaningful next step.

A concrete example is new engineers I've helped onboard. I tend to be useful when someone needs to turn an unfamiliar problem into a clear next step, especially when they are learning a new codebase and the team's operating practices at the same time.

At Lyft, I helped onboard two engineers who shipped production code in their first sprint. I gave them a specification template and a worked example, coached on-call triage and handoffs, and supported their implementation through review.

As they built context, I gave them room to own the decisions. One later drove a server-driven UI flow with minimal oversight. Their shipped work was theirs; the help I provided was structure, context, and support that they could eventually rely on less.

That's the kind of helpfulness I value: the person leaves with a clearer path and more ability to act. If someone is stuck, I want to understand the actual obstacle and help them move it forward.

In a few years, I want to be a technical lead who stays hands-on and owns larger projects from idea through launch.

I want to build that depth in health tech, understanding how the app, device, and user experience fit together. At Eight Sleep, I'd like to become someone the team trusts with its hardest Android problems, while helping other engineers grow.

I would trace Driver AI from the specification to the launched Android experience. The first boundary was the shared message model: another team owned it, and we needed Android and iOS behavior to align. I wrote the Android spec and worked on additive chat and tile models with that owner.

The implementation covered chat, speech input, plan cards, and driving-aware input behavior. That meant the Android client was responsible for more than displaying a server response; it had to handle the interaction in the context in which a driver would use it.

I also aligned client, backend, and data engineering on what data we were gathering, and coordinated with support and the driver team about our changes. This brought the teams responsible for data and the people supporting the driver experience into the work.

The next difficult boundary was integration. Duplicated messages and backend ownership gaps became launch blockers, and I helped work through them across teams. That is where I carried responsibility beyond a locally complete client.

The supported result is the public launch and later reuse of the chat patterns. The tradeoff I would emphasize is designing around shared contracts and existing consumers while delivering a new experience. The lesson is that end-to-end ownership includes making the system work across the responsibilities you own directly and the ones you have to coordinate.

A second project is the shared speech foundation and navigation voice choices. We had speech-to-text and text-to-speech work duplicated across Driver AI, navigation, and RiderAI, while drivers were asking for a different navigation voice.

I worked on consolidating the shared services and led Android integration across voice profiles, APIs, settings, and runtime behavior. The useful boundary was separating capability reuse from the product-specific choices, so the framework could support multiple experiences while navigation retained its own behavior.

I delivered through reviewable stages and also worked through provider throttling and cache warming. My large-PR feedback changed the way I structured later voice work: voice-to-voice used six pull requests and alternative voices used four. That made the review boundaries more explicit.

The framework was reused across three products. The reference reports more than 600 early switchers in roughly ten days of a staged release. That is initial uptake. I would keep the broader eligibility pool and any future audio monetization separate from the observed result. This project shows how I connect platform work to a concrete user need and carry the product integration through the runtime and release boundaries.

A third example is the AI delivery workflow I built at Lyft. It takes a product request through research and planning into draft pull requests for iOS and Android. The prototype made the idea visible, but the harder problem was making it something other engineers could operate responsibly.

I added a Redis-stream work queue, authenticated webhook ingress, and a GitHub App. I also added human review gates, addressed restart and cluster failures, and aligned the integration with the team responsible for the infrastructure. Those decisions established how a request entered the system, how work executed and recovered, and who accepted the generated change.

I packaged the workflow for broader use. The record reports sprint usage and another organization asking to adopt it. I treat those as usage and interest, with a quantitative productivity result still needing a defined cohort and measurement.

What I would emphasize is that I independently took a promising but ambiguous prototype and made the operating responsibilities clearer. That also gives me a useful connection to Andrew's article about bringing AI into real workflows: the practical value comes from context, integration, and clear boundaries around autonomy.

Yes. I understand the team is working closer to a 60-hour week right now, and I'm prepared for the workload you've described. I want to be part of building something special in a product area I care about.

The work I find most rewarding is hands-on ownership: getting a new product across the finish line, resolving difficult integration problems, and staying engaged when production needs attention. Driver AI and the Android Auto work are concrete examples of that.

I've also learned that moving quickly takes clear communication. On a voice project, I let scope growth and on-call work accumulate without making the capacity impact visible early enough. I owned the estimate slip and changed how I plan and communicate those tradeoffs.

I'd bring urgency, independent ownership, and that transparency. I'm interested in the responsibility that comes with helping this team build the next stage of the company.

A pressure example is a voice project where scope grew, backend gaps appeared, and I had overlapping on-call work. I absorbed those changes without making the effect on capacity visible early enough, and my estimate slipped.

I owned that with my manager and reset the plan. I started making capacity caveats explicit, raising scope changes sooner, and splitting later work into smaller reviewable units. My large-PR feedback reinforced that change in delivery structure.

What I learned is that intensity does not remove the need to communicate the tradeoff. I need to make the changed situation visible early enough for the team to adjust scope, sequencing, or ownership. The supported result is the replanning and changed behavior; I would give the actual delay and consequence from memory rather than invent a missed launch.

An Android Auto release failed when drivers connected to a physical head unit. The initial mitigation was broad because we had not yet isolated the affected population.

As a responder, I reproduced the failure on real hardware, compared phones and OS versions, and shared logs. I isolated the regression to Android 16, which let the team restore the feature for unaffected cohorts while validating the specific fix. I also helped check the release candidate and a manifest-test issue in CI.

The source reports a targeted fix within roughly three days. My contribution was reproduction, isolation, and validation; I was not the incident commander. The lesson was to distinguish immediate recovery from permanent remediation and to use evidence from the actual device and release boundary to narrow the risk.