Work

Axiom

A production database that replaced a $12,000-a-year vendor tool, unified member and donor records, and runs on about $120 a year.

Role
Designed, built, and shipped solo
Organization
Victory Life, three locations
Stack
Supabase, Postgres
Running cost
About $120 a year
$12k → $120
annual operating cost, before and after
API
into the marketing platforms the team already uses. The vendor had none
Built in
mail merge, exports, and a donor view the vendor never offered
Before: the vendor tool
SourcePlanning CenterCheck-in and attendance.
SourceDonor databaseGiving records.
Third-party glue$12,000 a yearJoined the two sources. No API. Lists took forever to compile.
OutputMarketing platformsNo API, so no connection.
OutputMail merge and exportsNot offered.
OutputDonor viewLeft unsolved.
After: everything flows into Axiom
SourcePlanning CenterCheck-in and attendance.
SourceDonor databaseGiving records.
Built and run in-houseAxiom on SupabaseOne production database for member and donor data. About $120 a year. Lists in a fraction of the time.
OutputMarketing platformsConnected over API to the tools the team already uses.
OutputMail merge and exportsBuilt in.
OutputDonor viewSolved, in the same database.
Vendor tool, per year$12,000
Axiom, per year$120
Same two sources in. Faster lists, an API into the marketing platforms, mail merge and exports, and the donor problem solved, out. At about one percent of the cost.

The problem

The organization was paying roughly $12,000 a year for a third-party product that joined its check-in and attendance system, Planning Center, to its donor database. It had three problems. It had no API, so nothing else could talk to it. Compiling a list took forever. And it cost $12,000 a year for that. It also never solved the donor side properly, and it offered no mail merge or export features. The renewal was coming up.

What I built

Rather than renew, I designed and built Axiom, a production database on Supabase. The same two sources flow into it, and it does the same job faster and cheaper, on infrastructure costing about $120 a year. Because it has an API, it connects to the marketing platforms the team already uses. It also solves the donor problem the vendor left open, and mail merge and exports are built in rather than missing. It is shaped around how the organization works instead of a vendor's assumptions.

I owned it end to end: spec, build, test, ship, and maintain in production. I built it solo, with a human-in-the-loop AI workflow doing the mechanical implementation while I owned correctness.

Why it matters

This is the clearest single example of a pattern I repeat. Find the expensive dependency, understand the real requirement underneath it, and build the leaner system that replaces it. The saving is the headline. The better fit to the organization's actual process is the part that lasts.

What it used