Founded and led development of a low-cost wearable device for CPR training, focused on accurately measuring compression depth and force to improve cardiac arrest outcomes.

Product ownership: Defined problem scope, success criteria, and feasibility boundaries for a medical hardware product under academic and clinical constraints.

Cross-domain decision making: Independently designed hardware architecture, circuit layout, and sensing pipeline; validated force accuracy (~2-inch / 125 lbs) through calibration and testing.

Expert-driven product validation: Proactively engaged medical school faculty and clinicians (via cold outreach) to validate assumptions, define product success metrics, and guide technical trade-offs.

User research & system thinking: Conducted user interviews and analyzed existing CPR training systems to inform interaction design and feedback mechanisms.

Status: First-generation prototype entered user testing and advanced to screening round of the Hollomon Health Innovation Challenge.

Led the end-to-end development and launch of PantryLink, a cloud-based IoT platform supporting 335+ micro-pantries across Washington State, enabling real-time visibility and community-driven donation coordination.

Defined product strategy, system architecture, and data flow for a distributed IoT system integrating hardware sensors (weight, door, temperature) with a web-based dashboard. Designed the cloud-based backend using Azure Functions and Cosmos DB to support scalable, real-time data processing across donations, wishlist, messaging, and telemetry streams.

Drove cross-functional execution across frontend, backend, and hardware teams under an aggressive delivery timeline. Established frontend–backend data contracts and normalization layers to ensure system reliability despite evolving backend schemas and hardware constraints.

Owned key product decisions around data governance and platform safety, including moderation strategy for user-generated content (donations, messages) and trade-offs between openness and control in a community-driven system.

Leveraged AI-assisted development tools (Cursor, GPT) to accelerate full-stack implementation, enabling a non-CS team to successfully deliver a production-ready system with cloud deployment and monitoring.

Owned MVP definition and product execution for a B2B2C AI-assisted property operations platform in a 0→1 environment, working closely with founder and engineering teams to translate ambiguous requirements into structured product decisions.

Product logic & system design: Transformed AI-generated concepts into production-ready product specifications, defining 20+ edge, empty, and failure states to resolve ~90% of critical workflow blockers and ensure system reliability.

Workflow & platform architecture: Designed multi-role workflows across property managers and vendors, covering assignment → scheduling → completion, and integrating AI-assisted decision support and communication flows.

AI product strategy: Defined the role of AI in project creation and vendor assignment, evolving the system from UI-driven inputs toward AI-assisted diagnostics and structured decision-making.

Cross-functional execution: Led weekly product iteration cycles with founder and engineers, aligning on product direction, resolving edge cases, and delivering complete happy paths with ~75% edge-case coverage.

Platform foundations: Established early product principles and system patterns to support scalability and future internal tooling.

Conducted 50+ user interviews and usability tests to identify trust and consent barriers in AI-driven financial workflows.

Delivered 3 trust-building UI interventions that reduced bank-linking drop-off by 24% and increased demo satisfaction by 13%.

Product impact: Insights directly informed fallback logic for >120s AI delays and helped scope MVP boundaries for responsible AI positioning.

Led UX design for a 0→1 Express KYC onboarding flow under strict compliance and backend constraints.

Product logic & feasibility shaping: Co-defined onboarding scope by balancing regulatory requirements, system dependencies, and user friction; mapped validation rules and 10+ edge cases impacting backend complexity.

Cross-functional negotiation: Worked closely with compliance and engineering teams to iterate on form logic, field dependencies, and progressive disclosure strategies.

Outcome: Contributed to a simplified onboarding experience that launched with significantly faster completion and improved approval outcomes.

This experience marked my first exposure to high-constraint product environments and highlighted the importance of backend awareness in product decision-making.