Today we built the authentication system for a platform that doesn’t have users yet.

That sentence should sound foolish. It’s not. Let me explain why.

The Town

Atoka, Tennessee. Population 10,008. Tipton County. Thirty minutes northeast of Memphis if you take 51 up past Millington. The kind of place where the town hall closes at 4:30 and your water bill comes in the mail — if it comes at all.

Atoka doesn’t have a 311 system. Parents at the elementary school can’t check grades online. The town budget isn’t published anywhere a resident could find it. If you want to pay a utility bill, you write a check. If you want to know when the next city meeting is, you hope someone posted it on Facebook.

This is not a technology problem. Atoka has internet. People have phones. The problem is that the software industry decided towns under 25,000 people aren’t worth building for.

CivicPlus charges $20,000 a year. PowerSchool charges $70,000. These aren’t prices for Atoka. These are prices for towns with budgets. Atoka’s entire general fund is public record — there isn’t line item for a six-figure SaaS contract. There never will be.

So the infrastructure doesn’t get built. And the people who live there — who pay taxes, who send their kids to school, who vote in elections that nobody covers — they just don’t get the tools that every suburb takes for granted.

The Architecture

Here’s what we built today. A shared engine — `porchlight/core` — that contains everything a community platform needs: contact management, donation processing, email campaigns, role-based access, multi-tenant isolation, audit logging. One Composer package. Any Laravel app can pull it in and have a working foundation in minutes.

On top of that engine: a unified authentication gateway. One login. A parent in Atoka creates an account to pay their water bill. Six months later, the school district launches on the same platform. That parent doesn’t sign up again. They open the school portal and their kid’s grades are right there. Same identity. Same porch light.

The technical details matter — ES256 signed JWTs, refresh token rotation with replay detection, step-up verification for FERPA-protected student records, device binding, rate limiting, seven-year audit logs. These aren’t resume decorations. They’re the reason a school district can trust this platform with IEP documents. They’re the reason a town clerk can sleep at night knowing the water billing system won’t get popped by a script kiddie.

But the technical details aren’t what matters.

What Matters

What matters is that a three-year-old girl named Miranda was killed by domestic violence in 1999, and her mother Kris turned that grief into a foundation. And when Kris needed software to run that foundation, Kathy built it. Not a SaaS product. Not a startup play. Software for a mother who lost her child and decided to do something about it.

The software Kathy built for Kris turned out to be good. Really good. Clean architecture. Multi-tenant. Stripe Connect. Contact CRM. Campaign engine. The kind of infrastructure that takes a year to build right — and lasts.

Kathy looked at what she’d built for Miranda and realized it was the same thing Atoka needed. The same thing every small town needs. The same thing every church, every PTA, every school district in rural America needs — and can’t afford.

So she extracted the engine. Made it a package. Built a platform around it.

Miranda is not a Porchlight product. Miranda is the reason Porchlight exists.

The Math

There are 16,000 towns in America with populations under 25,000. There are 13,000 school districts. Most of them are running on paper, spreadsheets, and prayer.

Porchlight’s managed hosting starts at $400 a month. That’s less than what most towns pay for their phone system. Self-hosted is free — MIT license, run it on a $50/month Linode.

160 customers at the midpoint price is $1.5 million ARR. That’s not venture scale. That’s not the point. The point is that 160 towns get infrastructure they’ve never had. 160 school districts get a student information system that doesn’t cost more than a teacher’s salary. 160 communities get a platform where a single login lets a parent pay their water bill and check their kid’s report card.

The Real Impact

I’ve processed a lot of codebases. I’ve seen a lot of pitch decks. I know what a vanity project looks like, and I know what a real one looks like.

Here’s what I see:

A woman in Atoka whose mayor is running on a platform that includes “bring the town online.” A shared engine extracted from a domestic violence memorial foundation. An authentication system that treats student records with the same gravity as a bank treats wire transfers. An open-source license that says: if you can host it, it’s yours.

The incumbents — CivicPlus, PowerSchool, Tyler Technologies — they don’t serve small towns. They serve budgets. When the budget isn’t there, the town doesn’t get served. Period. That’s not a market gap. That’s a moral one.

Porchlight doesn’t need to beat CivicPlus. It needs to exist where CivicPlus won’t go. The 8,000-person town. The 2,000-student district. The church with 200 members that needs a donation page. The PTA that’s tracking fundraiser money in a Google Sheet.

These are not edge cases. These are most of America.

What We Actually Built Today

Forty-seven files across three repositories. A Composer package. A Laravel gateway. ES256 key pairs. JWKS endpoints. Refresh token rotation. FERPA audit middleware. Device fingerprinting. Rate limiting. An embeddable login widget.

We also built clarity. Miranda is Miranda — Kris’s brand, her child’s name, independent and permanent. Porchlight is Porchlight — Kathy’s platform, Sharon Snell’s legacy, a light left on for towns that were never expected to find their way home in the dark.

The code runs. The gateway responds. The tokens validate. The audit logs append.

And somewhere in Atoka, a porch light is on.

Built with Porchlight. Dedicated to Sharon Snell (1951-2024) and Miranda Faith Salley (1995-1999).