Systems shipped, in production.

A young studio with real work to show. Each case below is something we actually built and shipped — no mockups, no vaporware. More on the way.

Case 08 / 2026 — Web App
Sector
Consumer Marketplace
Service
Custom Web Development
Timeline
Ongoing
Engagement
Founding product

A real marketplace for golfers, built from scratch.

Golfers buying and selling used clubs had no home of their own — just eBay and Facebook Marketplace, neither built for the way golfers actually trade.

VarsityGolf.org is a used-golf marketplace where real golfers buy and sell directly with each other, closing the gap the big generalist platforms left wide open. We designed and built the product end to end — listings, accounts, the two-sided buy-and-sell flow, and the infrastructure under it — and we continue to develop it as one of the studio's flagship products.

It gained real users quickly and remains one of our most exciting long-term projects — proof that we don't just build sites for clients, we build and run real products of our own.

Live & growing
in production with real golfers buying and selling
2-sided market
buyers and sellers in one purpose-built place
Built in-house
designed, engineered, and operated by the founders
Case 07 / 2026 — AI Integration
Sector
Commercial Real Estate
Service
AI Business Integration
Timeline
Ongoing
Engagement
AI Integration Partnership

Tenant intelligence automated for a boutique CRE advisory firm.

AI-powered tenant intelligence system for a boutique CRE advisory firm — dashboard and workflow overview

A 15-year retail CRE veteran was generating market intelligence the same way everyone else was — manually, backward-looking, and priced for institutions he no longer worked at.

We mapped his full workflow from tenant identification to deal outreach to client reporting and embedded AI at four integration points: automated tenant expansion signal monitoring by trade area and demographic fit, AI-structured outreach sequencing designed to start conversations rather than pitch, client reporting automation that compresses hours of intel compilation into minutes, and site selection analysis cross-referencing foot traffic, trade area demographics, and lease comps faster than manual CoStar pulls.

The first session delivered live improvements across both tenant research and client communication. The engagement continues on a retained basis — building out all four pillars at boutique scale, giving a solo advisory operation the intelligence infrastructure of a much larger shop.

4 pillars
research, outreach, reporting & site analysis automated
15+ yrs
of CRE expertise the system now scales
Day 1
live workflow improvements delivered in the first session
Case 06 / 2026 — Web Development
Sector
Political Campaign
Service
Custom Web Development
Timeline
Ongoing
Client
Name withheld

A working voter-engagement tool for a sitting judge's re-election.

Local campaigns run lean. The team needed a site that looked credible, captured supporters, and could change week to week — without a CRM or a developer on call.

We built and maintain a fast, multi-page campaign site on a lightweight serverless stack — no heavy CMS to pay for or babysit. Volunteer and endorsement forms write to a database and instantly email the campaign, so a new supporter never sits unseen in a dashboard nobody checks. A built-in one-tap contribution flow handles small-dollar donations from a phone, with the legally required disclosures applied automatically. Campaign-specific settings live in a single config block, so routine updates don't need a code deploy.

As an example of the ongoing work: endorsers can attach a photo with the candidate, compressed in their browser before upload and embedded straight into the campaign's notification email — optional and fault-tolerant, so an endorsement always goes through even if the photo doesn't.

3-in-1
volunteers, endorsements & donations in one place
0 CMS
serverless stack, editable without a developer
Real-time
every signup emailed straight to the campaign team

What could your number look like?