Creative ScientistWhat is a Creative Scientist?Find outWhat is aCreativeScientist?

I've worn many hats: Founder, Product Manager, Designer, Developer, and Research Scientist, which is why I call myself a Creative Scientist.

I can bridge every seat at the product table, taking ideas from research insight to design, code, and market-ready products used by hundreds of thousands.

This makes me a rare asset: accelerating product cycles, uncovering opportunities, and raising the bar for user experience.

& Entrepreneur

Companies and institutions

Research and academia

Products and systems

• Seena Labs' AI Ambient EthnographerSeena Labs' AI Ambient EthnographerAn ambient, always-on, AI-powered UX researcher that can be embedded in products with a single line of code.Learn moreVisit Seena LabsRead the launchSeena Labs is building AI products for understanding how people work with AI and software in general. Our products aim to understand the relationship and interactions between three entities: the people building intelligent software, the software itself, and the people using it.Our first product is the AI Ambient Ethnographer. Installed with a single line of code on software products, it observes and understands the user's journey and interviews the user during moments of interest throughout their interactions to provide a qualitative explanation to the quantitative behavior collected with usage analytics data.1 line of code install. Autonomous. Proactive insight generation. Contextual.
• 3rd Brain: AI Organizer3rd Brain:
  Couple's AI OrganizerAn AI Organizer for couples.Learn moreTry 3rd BrainAn AI organizer for couples: a mobile app and capture system that helped people turn thoughts, notes, and tasks into a more structured shared external brain. The work sat at the intersection of AI-assisted organization, mobile product design, mental load, and behavior change, and gave me direct founder experience building a consumer AI product from concept into shipped iOS and Android apps.What if couples had a 3rd Brain?When my wife and I were planning our wedding, we were both overwhelmed by scattered tools, fragmented thoughts, and invisible work. Sticky notes, shared docs, endless texts: none of it was enough. As two product leaders, we came up with an idea for a product, a third brain to help our two tired brains.But was it an us problem? Turns out, no. We started seeing patterns among our friends, dug deeper online, including reports from the UN, and found that mental load in relationships, especially among women, is a global, trillion-dollar problem. At the time, generative AI APIs were becoming ubiquitous, easy to call, and cheap. So we asked: What if AI could act as a third brain?A shared cognitive layer to reduce invisible labor and improve communication between partners.I got to work and built the MVP in three weeks around one hypothesis: use AI to reduce the load of data entry. By having users simply speak to the app, the AI should be able to identify tasks from the brain dump and assign them to the appropriate partner based on some user data we pass along. That is how we came up with the Brain Dump feature.We tested this prototype with 20 users, observing real-world behaviors in context. Seeing users' eyes widen when the app turned their messy brain dumps into structured tasks and notes shared with their partner in seconds was the green light we needed to build 3rd Brain.Talking to users, we noticed a constant theme: work is going unnoticed and unappreciated by both partners. So we made invisible work visible. We redesigned the app's home page to show tasks completed and notes added by the user's partner, and added the ability to send appreciation reactions. This was not about gamifying relationships. It was about validating effort and creating shared awareness.The next few months were a blur. During that time I:Built a waitlist through direct outreach and community events.Launched a 100-user closed beta, then opened up to the App Store.Hired and managed a developer and onboarded a contract GTM expert.Grew to 500 registered users over six months through organic acquisition.“This app is essential to my life now!” Diana, 43. New York.Shipped fast, iterated intelligently.Being the chief designer and user researcher, I stayed in constant contact with our users and even started offering free video calls to onboard couples onto the app. Along the way, I was able to:Double partner adoption through a redesigned onboarding flow that made partner invites feel essential, not optional.Reduce day-one churn by creating an onboarding email journey that guided users with behavioral nudges and practical examples.Identify key feature gaps, like integrated calendars, as high-priority.This was not just a productivity tool. It was an AI-native design exercise in emotional UX: about how people think, forget, remember, and relate.It required me to:Design interfaces that feel more like co-thinking than task apps.Build logic and UX atop generative AI.Think deeply about fairness, equity, and explainability inside the interface.
• Bilt Dining ProgramBilt Dining ProgramA restaurant rewards program that brought everyday dining into Bilt's loyalty ecosystem.Learn moreAt Bilt, I helped launch Bilt Dining across major metros by shaping the product experience, restaurant discovery surface, and rewards logic. The program connected neighborhood dining to the broader loyalty system and reached about 10% adoption in early rollout.10% user base adoption within the first 24 hours. And zero bugs reported.Press coverageArticle
• Bilt WalletBilt WalletA card-linking product that let non-Bilt cardholders earn rewards across everyday spend.Learn moreI helped introduce third-party card linking in Bilt Wallet by integrating transaction streaming APIs and designing the account-linking and rewards experience. The work expanded Bilt beyond the Bilt Mastercard and helped users connect multiple cards for rent-adjacent and neighborhood rewards.Press coverage
• Bilt x Lyft Rewards IntegrationsBilt x Lyft Rewards IntegrationsA partner rewards integration connecting Bilt members to Lyft account linking and mobility rewards.Learn moreI helped launch Bilt's Lyft partnership by defining API contracts, shaping in-app onboarding, and coordinating the consumer rewards experience. The integration drove 89K account linkages in the first 24 hours and connected rent, travel, and daily mobility into one loyalty loop.89,000 accounts linked in 24 hours. “This was the most thoroughly tested product launch in the history of the company.” - Bilt Head of Product during company-wide all hands meeting.Press coverageLyft Coverage
• Bilt x point.me Flight SearchBilt x point.me Flight Search PortalA flight rewards search portal that helped members compare points, miles, and cash.Learn moreI led the integration of point.me's flight rewards engine into Bilt's platform, redesigning the UX so members could make clearer miles-versus-cash booking decisions. The product became a high-intent travel surface with about 50K monthly searches.50,000 monthly searches.Press coverageArticle
• Amazon Low-Price Bundling ProgramAmazon Low-Price Bundling ProgramA grocery bundling program that helped customers find better value while improving shipping economics.Learn moreAt Amazon, I led product work for a low-price grocery bundling program across 12 markets. The program packaged small low-priced items into better-value baskets, helping customers discover savings while improving shipping efficiency and generating $55M in incremental sales and $120M in annual shipping savings.9 products launched. 12 global markets. $120M in savings. $55M in sales.
• Amazon Bundle Tracker and Promo SurfacesAmazon Bundle Tracker and Promo SurfacesShopping UX and promotion surfaces that made grocery bundle progress easier to discover.Learn moreI helped ship customer-facing bundle tracker and promotional surfaces across grocery shopping flows. The work included nine UX features and a U.S. promo page with roughly 100K daily visits, making an operational pricing strategy legible as a simple shopping experience.100,000 DAU. 5 global markets.
• Amazon Grocery Shopping ExperienceReimagining Grocery Shopping on Amazon.comA cross-org innovation workshop to reimagine Amazon's fragmented grocery shopping experience.Learn moreI led a week-long cross-org innovation workshop to reimagine the fragmented user experience of grocery shopping on Amazon.com.3 orgs. 25 participants from engineering, leadership, finance, and legal. 1 week. 20 user interviews. 5 product proposals presented to senior leadership.Amazon’s grocery ecosystem was deeply fragmented. Shoppers encountered disjointed experiences between Amazon Fresh, Amazon.com, and Whole Foods, each with its own workflows, interfaces, and fulfillment quirks. Despite having the infrastructure to serve millions of grocery customers, the experience failed to feel unified, intentional, or even intuitive.Leadership challenged our team to “Think Big” and reimagine the grocery experience at Amazon from the ground up.But the real challenge wasn’t just the problem. It was the scale and silos. Designers, PMs, engineers, and scientists all had different tools, processes, and communication styles. Innovation couldn’t happen in documents alone. It needed to be designed as an experience.As the Product Manager and Design Manager leading this initiative, I made a proposal: a three-day innovation offsite that blended Amazon’s “Working Backwards” with Design Thinking.I created a hybrid experience for 35 cross-functional stakeholders across design, engineering, science, and product. Some participants were in the room in New York, some in Seattle, and we even had a participant calling in from Pakistan.I built a bespoke process blending my background teaching Design Thinking classes at the University of Washington with Amazon’s PRFAQ framework. I crafted virtual whiteboards that guided participants to work in groups through multi-phased challenges:1. Pick from a list of problem statements I generated ahead of the workshop, informed by 20 user interviews with grocery shoppers that I conducted in partnership with the Amazon UX research team.2. Generate 100 ideas in 10 minutes.3. Pick a solution to assess problem-solution fit.4. Design how the user would use the proposed solution by telling a story.5. Communicate the idea using Amazon-native press release docs, PRFAQs.The collaborative and fun nature of the workshop created an inclusive space for technical team members to become storytellers, teaching engineers and scientists how to frame product ideas in user-first language.I participated in the workshop myself, contributing my own ideas while facilitating, guiding, clarifying, and giving feedback to participants, both in person and online.By the end of the third day, the team had eight PRFAQs written in three days, representing bold product proposals across personalization, grocery discovery, bundling, and fulfillment.Workshop outcomes fed directly into Amazon’s annual planning cycle, giving leadership tangible proposals ready for roadmap integration.The team also left with a repeatable innovation process, a template for future 0-to-1 product strategy efforts inside the org.This high-stakes project required me to facilitate vision without ego, translate user pain into product opportunities at scale, blend structured creativity with organizational literacy, and teach technical teams how to pitch like product thinkers.It’s a reflection of how I lead: designing not just the product, but the process that makes great products possible. Because Everything is designed.
• Itau Mobile Loan AppItau Mobile Loan AppA mobile loan-processing prototype for a major banking client through AWS.Learn moreThrough AWS product innovation work, I helped Itau explore and prototype a mobile loan-processing application. The engagement translated business goals, cloud capabilities, and customer needs into an experience direction the bank could use to modernize loan workflows, expanding the engagement from four pods to 12 and leaving the client with a reusable product-innovation methodology.
• FDA Approval Research PlatformFDA Approval Research PlatformA healthcare research platform concept for navigating FDA approval evidence.Learn moreOn an AWS healthcare engagement, I helped rescue and reshape a struggling FDA approval research platform by rebuilding stakeholder trust, introducing persona-driven design, and creating daily alignment rituals. The team delivered a validated redesign in six weeks and preserved a critical enterprise relationship.
• EV Charging Platform PrototypeEV Charging Platform PrototypeA fleet-charging product concept for electric trucks and industrial operations.Learn moreThrough AWS, I helped prototype a platform for electric truck fleet charging. The work connected operational constraints, driver and fleet-manager needs, energy optimization, and cloud architecture into a product story for managing charging infrastructure at scale, including a 60-second pitch video produced in 48 hours.
• Apple Vision Pro Multimodal Input InteractionApple Vision Pro Multimodal Input InteractionEarly multimodal interaction research for Apple Vision Pro using voice, gaze, and mid-air gestures.Learn moreAt Apple, I worked on early multimodal interaction research and prototyping for what became Apple Vision Pro. I led product design for a text-input interaction system that combined voice, gaze, and mid-air gestures, then ran co-design studies under secrecy constraints that challenged core assumptions and influenced the feature direction presented to senior AI/ML leadership.
• Distributed Interaction DesignDistributed Interaction DesignA thesis framework for running design research across distance, devices, and contexts.Learn moreDistributed Interaction Design was my Ph.D. thesis direction at the University of Washington. It used the web, machine learning, and human computation to let researchers and designers run elicitation, co-design, and sensemaking work with people across the world, a framing that became especially relevant during social distancing.PublicationsInteraction magazine ArticleFull Dissertation
• I am Iron Man: Designing VR Gestures Using Priming with Sci-Fi MoviesI am Iron ManA research study on how sci-fi primes the gestures people invent for VR.Learn moreThis project investigated how science-fiction imagery shapes the gestures people imagine for virtual reality. By using priming from movies like Iron Man, the study explored how cultural references can influence gesture elicitation and interaction design for emerging modalities.PublicationsCHI 2021 Article
• Anachronism by Design: Reimagining the Desktop Set of IconsAnachronism by DesignA study of how young adults understand inherited computer iconography.Learn moreAnachronism by Design examined how young adults interpret desktop icons whose meanings come from older physical metaphors. The work asks what happens when interface symbols outlive the objects that originally made them obvious, and how designers should rethink icon systems for new generations.Journal article
• Crowdlicit: A Two-Sided System for Running Remote Co-Design SessionsCrowdlicitA distributed co-design platform that scaled end-user elicitation studies online.Learn moreCrowdlicit was a two-sided research platform for conducting distributed end-user elicitation and design studies. It let researchers create studies, recruit participants remotely, collect participant-generated interaction ideas, and organize responses at a scale that would be hard to reach in a lab session.CHI paper
• Crowdsensus: ML Approach for Large-Scale Qualitative Data AnalysisCrowdsensusA machine-learning-assisted workflow for making sense of large qualitative datasets.Learn moreCrowdsensus combined human similarity judgments with unsupervised machine learning to analyze large sets of qualitative design responses. The system helped researchers find agreement patterns in end-user elicitation studies while keeping the human nuance of the data visible.UIST paper
• Dytective: A Game to Detect DyslexiaDytectiveAn award-winning game that used AI and player behavior to detect dyslexia.Learn moreAt Carnegie Mellon, I was one of four scientists who co-invented Dytective, a game that uses AI to detect dyslexia from how children play. I helped design and develop the game experience, translating linguistic error patterns into playful levels whose performance data could predict dyslexia risk with over 90% accuracy. The project won ASSETS Best Demo, placed seventh in the International AI & Robotics for Good competition, and went on to help more than 500,000 children through Change Dyslexia.CMU coverageASSETS demo
• Prototype to Evaluate How Searchers with Dyslexia Search the WebDyslexia Search PrototypeA Microsoft Research prototype for studying search accessibility and dyslexia.Learn moreAt Microsoft Research, I worked on accessibility and dyslexia-focused search research with the Ability group. Building on interviews with people with dyslexia, I ran large surveys, developed a web-based prototype for rating page readability, and helped publish work that informed how search systems could better serve people looking for readable, useful results.Microsoft blogCHI paper
• ML-Based Self-Adapting Accessible InterfacesML-Based Self-Adapting Accessible InterfacesMachine-learning interfaces that adapt to changing pointing performance and ability.Learn moreAt UMBC, I worked as an accessibility researcher while completing my master's in Human-Centered Computing and bachelor's in Information Systems. This project explored adaptive interfaces that respond to user ability, motor performance, and changing needs rather than forcing every person into a fixed interface.W4A paper
• Mobile Gesture Authentication SystemMobile Gesture Authentication SystemA gesture-based authentication system for mobile devices and accessible input.Learn moreThis UMBC research explored mobile gesture authentication as an alternative to conventional passwords. The project investigated how gestures, touch, and mobile sensors could support secure interaction while opening questions about accessibility, memorability, and input performance.
• Sequential Gestural Passcodes Using Google GlassSequential Gestural Passcodes Using Google GlassA wearable authentication project using gesture sequences on Google Glass.Learn moreSequential Gestural Passcodes extended my gesture-authentication work into wearable computing. Using Google Glass, the project explored how a sequence of gestures could become a passcode system for emerging devices where keyboards and traditional authentication flows do not fit.
• InsterpreterInsterpreterMy first software project, inspired by working as an interpreter at Johns Hopkins.Learn moreWhile I was in college, I worked as an interpreter at Johns Hopkins. That job inspired me to build Insterpreter, my first piece of software: an early attempt to turn a lived communication problem into a tool that could help people coordinate interpretation work more effectively.