About
I want to do the highest-impact work there is — and I've spent a career learning how.
The work
Frontier technology, three times over
Across nearly twelve years at Google I kept gravitating to the same kind of problem: take technology at the edge of what's possible — a brand-new product category, AI that has to run on a phone — and turn it into something real people use and love. The win is never the technology for its own sake. It's the product, and the lives it changes.
I did it for families as the founding engineering leader of Family Link, growing it from a prototype into a product hundreds of millions of parents trust to help their kids grow up well online. I did it for AI by leading a broad portfolio across Google's Assistant and Gemini stack — device to cloud, product to infrastructure — as one of the pillars of a 150-person org. And I did it at YouTube scale, setting the direction for the mobile apps two billion people watch on. Different domains, one instinct: start from the person, and let the technology earn its place.
How I lead
Product instinct, engineering rigor, written down
I'm a well-rounded leader by choice: close enough to the code to earn engineers' trust, close enough to the user to know what's worth building, and able to carry both into the room where the company decides. I lead with documents — charters, one-pagers, roadmaps — because a team that can read the strategy doesn't have to guess it. Every team I've built or rebuilt started life as a written argument someone could push back on.
- Teams from scratch
- Hired and managed 25–40+ person organizations through ambiguous 0→1 platform charters, across Zürich, Mountain View, and LA — managing managers, including two EMs. One early team grew from 3 junior engineers to 8, with four promotions on my watch.
- Fair evaluation
- Authored calibration-session practices used beyond my own teams: time-capped discussions, bias pre-reads, and a standing rule — for every rating, ask why not higher and why not lower.
- Honest transitions
- Led teams through layoffs, strategy pivots, and re-orgs — including winding down scope my own teams built. Teams forgive a changed strategy; they don't forgive a hidden one.
- Users at the center
- Mission-driven culture grounded in who the system serves — the parent at midnight, the teen negotiating autonomy, the nurse double-checking an order.
Now
The founder chapter: AI for healthcare
I began building a healthcare AI company in late 2025 and left Google in early 2026 to go all-in. It's in stealth, built hands-on with early clinical partners. I chose healthcare because it's the highest-impact place I could take what I do best: frontier AI, aimed at a problem where getting it right changes people's lives, and getting it wrong isn't an option.
Founding means owning all of it — the product, the engineering, the customers, the hiring. I sit with clinical teams to figure out where AI genuinely helps, what it should do on its own, and when it has to defer to a person. It's the same instinct that ran through everything before it: powerful technology is only worth shipping when it earns the trust of the people who depend on it.
Education
Trained as an engineer, schooled by scale
- Stanford Graduate School of BusinessGraduate coursework, 2020–2022
- Santa Clara UniversityM.S., Computer Science, 2014
- JNTU HyderabadB.Tech., Computer Science, 2012
The rest of the curriculum was Google: COPPA and GDPR turned into working code, privacy review for 25+ launches, access-control design with the team behind Google's authorization systems, and years of learning that the systems people trust are the ones that show their work.
The long story
Twelve years, two rails: the platform, and the job of leading it.
- 2014–2016
Exploring the space
Google asks what it owes families
Kids were using every Google product, but nothing was built for them or their parents. A new Kids & Families effort ran parallel experiments — family location sharing, supervised devices, accounts kids could actually have — to find what families genuinely needed.
Founding engineer, Kids & Families
One of the first engineers on Google's family effort. Built the family location product that proved real demand — the first team at Google to benchmark live location — and ran the trusted-tester program for the area's sensitive early launches.
- 2016–2017
Convergence & launch
Four prototypes become Family Link
Google's answer to a worry most parents shared: how do I help my kid navigate technology? Four experiments merged into one product — set screen-time limits, approve apps, see location, lock devices at bedtime. It launched in 2017 as the first way for kids under 13 to have a Google account their parents managed.
Engineering lead, Family Link launch
Owned the launch engineering end to end — Android, iOS, web, and backend. Turned a set of prototypes into a shipping product, and into the foundation Google's family experiences would build on for a decade.
- 2018–2019
From app to platform
Supervision for teens and existing accounts
The product's most-requested leap: let parents supervise the accounts their kids already had — including teens — across phones, Chromebooks, and Google services worldwide. The hard part was doing it with genuine consent on both sides, never coercion.
Led supervision platform engineering
Made the case to grow Family Link from a kids' app into a system for whole families, then led the design — the parent–teen consent handshake, the account model, region-specific behavior. A small team I led shipped the core in seven months against 12–18 month estimates.
- 2019–2020
Earning back trust
Fixing the foundation
Fast growth strained the basics: ratings slipped, support lines filled, and parents — who need this product to just work — started to doubt it. The team paused new features and rebuilt reliability until the product deserved the families using it.
Led the reliability turnaround
Persuaded executives, many of them Family Link parents, to pause the roadmap — then directed the ~25-person, cross-site program. Complaints fell ~70% while the user base grew ~500%.
- 2020–2022
Digital wellbeing
From parental controls to family wellbeing
Family Link became the platform behind Google's youth experiences — supervision, family accounts, healthy-habit tools — used by products across Google. A full redesign and rewrite carried it past 200M people, and it set the bar competitors answered to.
Engineering leader, Family Link
Grew the engineering organization to 25+ while steering a 90-person cross-functional org. Owned the modernization bets, the executive narrative, and the product's standing with families and regulators worldwide.
- 2022–2024
The Gemini era
Google reinvents its assistant around Gemini
As Google pivoted into the Gemini era, its assistant had to be rebuilt end to end — natural conversation in Gemini Live, fulfillment that takes real actions, on-device intelligence for speed and privacy, and a seamless handoff between phone and cloud — then the two parallel stacks unified into one, for a billion users.
Engineering leader across the Assistant–Gemini stack
Led a broad portfolio — device to cloud, product to infrastructure — by managing managers and senior staff, and was one of the pillars of a ~150-person org, helping run its operating rhythm and performance calibration through layoffs and the Gemini pivot.
- 2024–2026
Two-billion-user scale
YouTube's mobile foundation, modernized
The first major modernization of YouTube's mobile architecture since 2016 — faster, more reliable, easier to build on — so hundreds of engineers could ship more for the two billion people who watch on mobile.
Platform strategy lead, YouTube mobile
Authored the multi-year strategy and won its adoption across four apps and the hundreds of engineers who build on them — including an AI-assisted approach to large-scale migration.
- 2025–now
Founder
AI for healthcare
The highest-impact place I could take everything I've learned about shipping frontier technology responsibly: clinical AI that helps real patients and clinicians, with humans in the loop where it counts.
Founder — product & engineering
Building the company end to end with clinical partners: the product, the team, and the judgment for what AI should do on its own and when it must defer to a person.
If you're building something where trust is the product — I'd enjoy talking.