This is the clearest proof inside the project that Genny can do more than talk about AI. A Buffalo-area clinic had an operational software fit problem. The answer was not more theory. It was a new system designed around the business.
Named with consent. Metrics are still being gathered, so this page focuses on the business problem, the build, and what it demonstrates.
The stack
React + TypeScript, Supabase, TanStack Query, Tailwind, shadcn/ui, and Clover, built as an AI co-developed project used day-to-day for interface work, schema design, and implementation.
Balanced Infusions is a Buffalo-area IV therapy clinic. The team was running on Vagaro, an off-the-shelf system that covered basic scheduling but kept forcing the operation into workarounds.
Appointment logic, intake, clinical notes, staff coordination, and reporting did not fit the way the clinic actually worked. The software was shaping the business instead of supporting it.
Most businesses do not need a custom application. They need better use of the tools they already have. Balanced Infusions was different because the mismatch was operational, recurring, and expensive.
Replacing one generic platform with another would have repeated the same problem. The better move was to design a system around the actual workflow and the people using it.
The product stack includes React, TypeScript, Supabase, TanStack Query, Tailwind, shadcn/ui, and Clover for payments.
AI was used as a co-developer throughout the build: for interface work, schema design, policies, and implementation support. This is one of the clearest proofs inside the project that Genny can move past advice into real delivery.
The app covers scheduling tailored to IV therapy workflows, integrated SOAP notes, intake built for the actual front-desk process, staff coordination, payments, and reporting shaped around what the clinic needs to see.
The point was not just feature parity. The point was operational fit: fewer workarounds, cleaner ownership, and a system that matches the business instead of fighting it.
The app, admin tooling, and documentation were built for handoff. The long-term goal is not to keep the owner tied to a developer for ordinary operational changes.
That same principle shows up everywhere in the new Genny model: build something useful, teach enough for the team to own it, and avoid creating dependency theater.
The business side starts with a free AI assessment because most companies need clarity first. When the right next move is a real implementation, this case study is the proof that the work can actually get done.
That does not mean every company should commission a custom application. It means Genny can help judge when the answer is training, when it is workflow redesign, and when it is a real build.
What to take from this
The real lesson is not “everyone should build software.” The lesson is that the next step should match the problem. Sometimes that is a workshop. Sometimes it is process redesign. Sometimes it is a tighter implementation sprint. Very occasionally, it is a real custom build.
The practical takeaway
The assessment is the right front door because it helps decide whether the answer is education, implementation, redesign, or simply waiting. This case study exists to show that when the answer is real build work, there is actual delivery capability behind the site copy.
Leave your email and we will send the fuller case-study update as the operating metrics are finalized.