Hi, I'm James.
I'm a product manager, and the work I'm best at is the complicated kind: sitting with customers until the real problem surfaces, then taking it from nothing to shipped. The last few years have been early-stage startups and tech consulting, where things move fast and I own the whole thing. I studied electrical engineering and computer science at Columbia and started my career as an engineer, so I'll read the code to work out why something broke, spend a day in Figma, or write the go-to-market plan myself. I've co-founded a company and built platforms from scratch, and I'm always picking up a new skill or building something on the side, usually tinkering with AI.
On AI
PM is turning into product engineering.
AI is redefining what the job is. Orgs are flattening, one person can carry a problem end to end, teams ship quicker and align in hours instead of weeks. Building the systems that make that true is the work now, not writing the document that describes it. It cuts the other way too. AI is a solution, not the solution, and plenty of the time the honest answer is that it doesn't move the outcome at all. Working out where it actually earns its place is the judgment call, and I'd rather ship the boring thing that works than the impressive thing that doesn't.
Where I've worked
Stringbean↗Lead Product Manager
Valley Stream, NY/Feb 2026 to Jun 2026/Contract
A proptech startup. It gives property owners and managers real visibility into maintenance operations, and holds the work accountable.
- First PM in the company. Built the roadmap and the prioritization framework, ran execution across engineering, design, sales and customer success, and shipped the first release in four weeks.
- Redesigned vendor management and shipped a Slack integration, scoped from 12 customer interviews and piloted with the largest account before it went out to everyone.
- Wrote the first customer documentation and built the self-serve features behind it, which cut support tickets by 70%.
- Prototyped whole flows with Claude and designed the UI in Figma, taking design-to-spec from days to hours.
- Read the usage data in SQL and sunset the features nobody used, which consolidated the navigation and put the roadmap back on the workflows customers actually ran.
Qwohter↗Co-Founder
New York, NY/Apr 2025 to Jan 2026
A unified workflow platform for dealer and distributor businesses. Project management, quoting and billing in one place.
- Built a B2B SaaS platform with AI-assisted quoting and operational workflows, which took 75% off quote-to-delivery and halved manual errors.
- Ran discovery with prospects until quoting turned out to be the worst of the friction, then scoped the first build around quote-to-delivery before adding project management and billing.
- Worked with three paying pilot customers to sharpen usability and the core workflows, and quotes generated per user went up 2.5x.
- Wrote the go-to-market playbook and ran it myself: demos, sales and onboarding end to end for the first accounts, in a shape the next person could repeat.
Leap Consulting Group↗Product Manager, Technical Delivery
Elmwood Park, NJ/Jul 2024 to Jun 2025/Full-time
A tech consulting firm. It builds solutions for telecommunications and healthcare companies.
- Led delivery of a $1M+ captioned calling app on Android for the firm's largest enterprise client, owning timeline, scope and execution across engineering and the client's stakeholders.
- Shipped the full build on schedule for a planned 750,000+ device rollout, through mid-project scope changes.
- Ran a remote team of four engineers on daily standups and sprint planning, and took 20% off dev cycle time.
- Worked with client stakeholders on UI redesigns and feature expansions, writing the acceptance criteria and rollout plans that lifted adoption by 20%.
- Built automated testing pipelines and tightened QA, cutting manual QA by 15% and getting defects found the same day.
Tekno-RoboticsProduct Engineer
Nanuet, NY/2023 to 2024/Full-time
A robotics startup. It builds a smart vending machine that pours nitro cold brew coffee.
- Worked directly with the CEO on where the machine went next, which put me in the product decisions rather than downstream of them.
- Designed proprietary, patent-pending parts in SOLIDWORKS and made them in house, saving over $100K.
- Designed 60+ precision parts in CAD, picking the materials and processes as I went, and took 30% off the cost of the machine and 12% off its weight.
- Built the manufacturing strategy, including the material and production tracking software behind it, which more than halved time to production.
Lilker Consulting Associates↗Electrical Engineering Intern
New York, NY/May 2022 to Jul 2022/Internship
A consulting firm providing mechanical, electrical and plumbing engineering services. The firm is IMEG now.
- Designed the electrical systems for a three floor school in AutoCAD MEP, and got the drawings finished in 25% less time than planned.
- Worked with the mechanical and plumbing engineers to place the electromechanical devices, which moved design approval 15% faster.
Education
B.S. Electrical Engineering
Minor in Computer Science
Columbia University
School of Engineering & Applied Science
Class of 2023 · New York City, NY
Second phase of the dual degree program with Yeshiva University.
B.A. Physics
Yeshiva University
New York City, NY
First phase of the dual degree program with Columbia.
Skills & Tools
Product
PMThe day-to-day.
Discovery & User Research
Confluence · Notion
Roadmap & Prioritization
Jira · Wrike · ClickUp
Specs & Decision Records
Confluence · Notion · Google Docs
Analytics & Experimentation
PostHog · Sheets
Release Monitoring
Sentry
Design & Prototyping
Figma · Framer · Claude · ChatGPT
AI in the Loop
Granola · Fireflies · Wispr Flow · Higgsfield
Cross-Functional Collaboration
Slack · Teams · Google Workspace
Software
SWWhat I build with.
Python
Django
Java
Spring Boot
TypeScript & React
Next.js · this site
APIs
REST · JSON · Postman
End-to-End Testing
Playwright
AI-Assisted Development
Claude Code · VS Code
Data & Platform
INFWhere things live and how they ship.
Relational Data
PostgreSQL · MySQL
Managed Backends
Supabase
Document Stores
MongoDB
Version Control
Git · GitHub · Bitbucket
CI & Deploys
Vercel
What I actually believe
Cutting is the hard part.
Getting it to work is the easy half. Knowing what to leave out is worth more than building the thing you guessed people wanted.
Outcome focused.
The measure isn't how much shipped, it's what changed for the person using it. AI is the easy example: it gets added to everything, and most of the time nothing about the job gets better. If a dropdown does it better, ship the dropdown.
I'd rather ship it and find out.
I've spent weeks planning things that turned out to be wrong. Two days of people actually using it would have told me.
Good design is key.
It's what makes a product easy to pick up and obvious to use. It's also what makes someone want to keep using it, and that part is a feeling, not a feature list.
Customers know the problem,
not the solution.
People ask for a feature. Take it seriously, then keep asking why until you hit the thing underneath. That is the part worth building, and it usually fixes it for everyone else too.
Automate the boring parts.
If someone on the team does the same thing by hand every week, that is a job for a script or a tool. I go looking for those. Hours back is the cheapest win a team ever gets.
Some ideas I've had
I've explored ideas in autotech, consumer, founder tools and retention. Some came out of a problem I kept running into myself. The rest are problems I watched other people put up with, where the fix seemed obvious enough to be worth writing down.
Eazale. Leasing a new car is a day of paperwork and a monthly number nobody will explain. The whole process should run itself.
Sohde. Dating apps put two strangers in a bar with nothing to do but audition each other. Meet through the thing instead: pickleball, a wine tasting, whatever you were already going to do.
Backboard. The earliest stage of a company is the one where nobody is available to tell you the idea is wrong. Crowdsource feedback and validation from other founders sitting at the same point.
Offboard. A cancel-flow dropdown gives you "too expensive" and nothing else, so retention offers get built on a guess. An AI exit interview inside the cancel flow that gets at why someone is really leaving, and turns that into churn data good enough to build the offers on.
Off the clock
Hobbies
Sports, most of them
Basketball, baseball, tennis, golf, pickleball, skiing, fishing.
Yankees, Giants, Knicks, Rangers
New York, all four. Watching at home is fine, but the thrill only really lands when you're in the building.
CrossFit
Best training I've found. The strength carries into every other sport.
DJing
EDM mixes.
Investing
Public markets, mostly. I want to understand a business well enough to say the thesis out loud in two sentences.
Coin collecting
Old Americana, inherited from my father. The dates on them go back further than anything else I own.
Reading
Startups, AI, product, and entrepreneurship.
In my orbit
