Problems worth solving — and the thinking behind them
Every award and patent here started as a question: why does this problem still exist? The recognition came later. The question came first. These are the stories of how we got from one to the other.
Two ideas that took years to prove
Privacy Exposure Risk: Making the Invisible Measurable
The internet accumulates information about people in ways they rarely see and almost never control. Data brokers, aggregators, and old accounts quietly build profiles — home addresses, employment histories, family relationships — available to anyone who looks. Most privacy products respond to this reactively: they clean up data after it's already been exposed.
The question that drove this patent was a different one: what if users could know their exposure risk before something went wrong? What if the system was predictive rather than reactive?
The invention we built analyzes an individual's digital footprint across online entities — public records, data broker databases, web presence — computes a meaningful risk index, and delivers it in a way that's actionable rather than overwhelming. The hardest part wasn't the NLP pipelines or the data aggregation. It was deciding what "risk" actually meant for a non-technical user, and building a scoring system that was honest about uncertainty rather than falsely precise.
This patent represents a shift in how I think about security products generally: the most valuable thing you can give a user isn't a list of what's been cleaned up. It's a clear-eyed picture of where they stand — and a path forward.
Digital Exhaust: The Problem Nobody Was Solving
Most people who use the internet actively have dozens — sometimes hundreds — of accounts they no longer access. These dormant accounts represent a specific class of privacy risk: they contain personal data, they're often forgotten, and because no one is checking them, they're vulnerable to breach without the owner ever knowing.
We called this "digital exhaust" — the trail of stale accounts and abandoned services that accumulates over a lifetime of internet use. The problem had been acknowledged for years. What was missing was a systematic method for identifying it and doing something about it.
The invention analyzes a user's digital footprint to identify dormant accounts and quantify the risk they represent, then automates or guides the process of reclaiming, removing, or securing them. The technical challenge was in defining what "stale" meant across wildly different platform types — an abandoned social media account carries different risk than a forgotten financial services login. The system had to be sensitive to that distinction.
This was privacy engineering at its most practical: not preventing new exposures, but systematically reducing the surface area of old ones. Less glamorous than building new capability. More impactful for most users' actual risk profile.
The moments that changed how I think
Selected from 8+ global awards across 15 years. Each one is here because it represented a genuine insight — not just a deliverable.
From Idea to Product in 48 Hours — and What That Proved
The 2024 SurveyMonkey Global Hackathon posed a question that market research platforms had been circling for years: could you capture real-time industry trends and segment them by meaningful client characteristics — in a single, coherent product experience?
The team I led built exactly that. A platform that pulled live industry data, applied advanced segmentation, and surfaced actionable insights in a format that business users — not data scientists — could work with. Selected from global entries, the win mattered less than what it demonstrated: that a team with real creative energy, given the space to work without process constraints, will consistently exceed what a roadmap would have produced.
The more important outcome was strategic: parts of the hackathon prototype found their way into product development. When innovation and shipping can be connected, the whole organization benefits.
Designing Privacy Infrastructure Before GDPR Made It Mandatory
In 2016, GDPR was two years away from enforcement. Most organizations in the consumer technology space were treating personal data protection as a legal problem — something to be handled by compliance teams at the point of regulatory requirement, not an engineering discipline.
I proposed and prototyped an AI-based personal data protection system that treated privacy as an engineering concern: systematic, automated, integrated into the product from the beginning rather than bolted on afterward. The proposal influenced Intel's product roadmap — and years later, when GDPR changed the landscape, teams that had built privacy into their architecture were in a dramatically better position than those who hadn't.
The award was recognition for the idea. What I remember is the lesson: the best time to build for a requirement is before it's required. The organizations that lead on privacy, security, and compliance are almost always the ones that treated them as product values rather than regulatory checkboxes.
Stolen Device Recovery: The Feature That Started a Product Line
Device theft was a significant problem in emerging markets — and existing anti-theft solutions largely relied on user-initiated recovery flows that required the victim to know their device had been stolen and act immediately. By the time most users realized what had happened, the device was wiped and untraceable.
The idea I developed — Media Capture for Stolen Devices — flipped the model. Rather than waiting for the user to respond, the system would autonomously capture multimedia evidence (camera, location, usage patterns) the moment specific theft-indicating behaviors were detected, and transmit it silently to a recovery service. The evidence would surface to the user and, where applicable, to law enforcement.
The insight was behavioral: a thief using a stolen device behaves differently than the legitimate owner. Detecting that behavioral delta, rather than relying on explicit device-level signals, made the system substantially more reliable. That thinking — modeling adversarial behavior to build better defensive systems — influenced my approach to security product design for years afterward.
GPS-Free Location: Rethinking a Hardware Dependency
In 2009, GPS was the default assumption for location-based services — but GPS hardware added cost, battery drain, and acquisition latency that made it impractical for many emerging-market use cases. The question was whether location-based services could be delivered meaningfully without hardware-level GPS.
The solution we developed used a combination of cell tower triangulation, Wi-Fi positioning, and behavioral inference to deliver location accuracy sufficient for a wide range of consumer LBS applications — without GPS hardware dependency. Winning the Asia-Pacific NAVTEQ challenge validated both the technical approach and its market relevance.
More importantly, it instilled an engineering habit that I've carried ever since: before accepting a hardware or infrastructure dependency as a given, ask whether the actual user need can be met a different way. The constraints of emerging markets are a remarkable teacher in lean, assumption-free design.
Additional recognition includes: Tech Talk Winner (McAfee, 2023 and 2022) for presentations on AI-driven privacy protection and cross-platform social media privacy management; Technical Excellence and Strategy Award (McAfee, 2022) for delivering a compliance-grade privacy platform at million-user scale; Employee of the Quarter (Intel, 2018) for improving service response metrics through architecture and process optimization; and ACE Excellence Award (Intel, 2014).
These are the records of work done well enough to be noticed. The work I'm proudest of is harder to put in a list.