I help early-stage founders turn a strong vision into a product customers want and their team can build. Through Align Product Studio, I help teams get clear on the target customer and strongest opportunity, test the assumptions that matter, and translate what we learn into an evidence-backed product direction, prototype, and build-ready scope.
The work below shows how I use prototypes and lightweight builds in my product leadership work: to make decisions tangible, test assumptions early, align teams, and solve real workflow problems without overbuilding.
Client repositories are private, so I've focused on the product problem, the decisions I helped make, and what the work enabled.
Want to talk through a product challenge? Book a call.
The examples below show how I work across different stages of a product: clarifying the product strategy, bringing an idea to life, and helping an existing product evolve.
| Product challenge | What I did + what it enabled |
|---|---|
| Narrowing four possible markets into one focused zoning product | Helped narrow four potential target customers to one, tested the direction with prospective buyers, and translated what we learned into a prototype, focused scope, and build-ready documentation. I stayed involved through the build as early feedback shaped the next iterations. |
| Choosing what to build next from four competing product ideas | Worked with the CEO and COO to get several ideas out of their heads and into rapid prototypes they could react to and compare. Leadership aligned on one priority, and the development team had enough clarity to take the chosen direction and run with it. |
| Making a complex retrofit-planning vision tangible | Worked with the CEO, CPO, and CTO to translate a PowerPoint and broad vision into a working prototype using live mapping and property data. The team now uses it in client demos, investor meetings, and partner conversations, and it is detailed enough to serve as a starting point for development handoff. |
| Designing a scalable support flow from a messy customer-service process | Discovery with the team surfaced that customer emails, calls, and unclear ownership were consuming significant time. I designed a guided experience that helps users self-serve where possible and routes more complex requests to the right team, then delivered the prototype and PRD for implementation. |
| Product challenge | What I did + what it enabled |
|---|---|
| Building an insurance product from an early-stage idea | Defined the product, workflows, and requirements, built the first prototype, then worked with an engineer to build a functioning application. It helps small business owners identify potential coverage gaps, understand renewal changes, and know what questions to ask before renewing. |
| Reframing a community website as a self-service operating platform | Redesigned the experience around a few clear resident actions, then built the live site and an admin platform so non-technical committee members can manage memberships, events, newsletters, content, and resident requests themselves. View the live site → |
| Product challenge | What I did + what it enabled |
|---|---|
| Evolving a live coaching platform without losing product coherence | Reviewed the B2B and B2C experiences, identified and prioritized product gaps, prototyped uncertain feature ideas before development, and supported a broader rebuild with product requirements and QA. The result was a more intentional product experience and clearer roadmap for the development team. |
| Building a repeatable outbound workflow from a defined ICP | After defining the ICP, I expanded an initial request for an Apollo lead list into a lightweight workflow that researches accounts and contacts, identifies the strongest-fit leads, and drafts personalized outreach. It gives the team a prioritized first lead list without hours of manual research. |
I also build products around problems I find interesting. It keeps me close to the craft and gives me a place to test new product ideas, AI capabilities, and interaction patterns without waiting for a client brief.
Built while my co-op board was hiring a building superintendent. Previous candidates had been evaluated inconsistently, with no durable record of why someone advanced or was rejected.
I designed a workflow that filters hard requirements upfront, evaluates every resume against the same building-specific criteria, and keeps interviews, references, and scoring together. AI supports the evaluation, but the board makes the decision.
We used it throughout the hiring process, creating a much more structured approach. Our property manager saw enough value in it that she wanted to use it for other buildings she manages.
Built with: React · Vite · Firebase · Claude API
A privacy-first product for noticing patterns in your own behavior.
It looks for repetition in transaction and location data, then surfaces observations without turning them into scores, goals, or judgments. Those constraints are intentional: the goal is awareness, not optimization.
Privacy is part of the product architecture, not just the policy. Raw location coordinates are discarded during processing rather than stored for later use.
Currently in active development.
Built with: Python · FastAPI · PostgreSQL · React · TypeScript
A French learning product I originally built for one very specific user: my dad, an advanced learner working toward C1 proficiency.
Rather than build another general language app, I focused on the gaps he actually had: pronunciation feedback, more challenging conversation, and culturally relevant material. The product combines real-time pronunciation analysis with authentic French content and targeted practice.
He eventually shared it with his French class, giving me a small but useful set of real users to learn from.
Built with: React · TypeScript · Claude API · Whisper · HuggingFace Transformers · WaveSurfer.js
NYC recycling rules are surprisingly difficult to apply in the moment. I wanted to test whether computer vision could remove that friction: take a photo of an item and get an immediate, NYC-specific answer about what to do with it.
I built the first version in four days to see whether the interaction was useful enough to make a confusing civic process feel simple.
Built with: JavaScript · Gemini Vision API · Netlify
I use whatever gets us to a useful answer fastest. Most of my work is built with modern web tools and AI APIs, with an emphasis on speed, clarity, and avoiding unnecessary complexity.
Frontend: React · TypeScript · Next.js · Vite · Tailwind
Backend & data: Python · FastAPI · PostgreSQL · Supabase · Firebase
AI & integrations: Claude · Gemini · Whisper · HuggingFace · Google Places · ArcGIS · Tavily
Deployment: Netlify · Vercel