How FlagStat Was Vibe-Coded with Bolt: Stack, Pricing and First Users

FlagStat went from one prompt to the official scorekeeping app of Florida's state flag football playoffs. Its builder, Bradley Metrock, describes himself as not an engineer. He built the web app on Bolt.new, an AI tool that writes and runs full web apps from chat prompts.

Revenue is not public. But the traction is: the homepage now claims more than 1,500 coaches across 30+ states. It also says FlagStat is the official and exclusive flag football stats platform of the NFHS, the body that writes US high school sports rules.

The core point: a non-developer's biggest edge was knowing the sport's rules and people, not the code.

His background

Metrock runs Project Voice, a podcast, conference and venture firm focused on voice AI. So he is not new to tech. He has spent about a decade around conversational AI.

But he is not a software engineer. The Bolt story makes that plain, and he frames himself as someone who simply found the problem interesting.

What he did have was a clear niche. The Bolt story calls FlagStat purpose-built for flag football, not a general sports tool. It supports three rule sets out of the box: NFL FLAG, NFHS and NIRSA (college intramurals).

Tool choice

He started on Bolt.new. The first prompt was roughly "build a flag football scorekeeping application." Then came iteration. In his words: "Bolt reduced the complexity of the problem down to 30 or 40 more prompts after that."

For phones, he used a different AI tool that specializes in native mobile apps. The Bolt story does not name it. That app, FlagStat Go, runs on iOS and Android.

The glue is Supabase, a hosted database with login built in. Both the web app and the phone app talk to the same Supabase backend. That is a sensible pattern for non-developers. One database means one source of truth, even if two different AI tools built the front ends.

Where he got stuck, and what he paid for

Here the public record is thin. The Bolt story does not describe specific roadblocks. It does not mention paid developers or contractors either.

What it does describe is a working rhythm. Coaches send feature requests on Thursday. A new release ships the following Tuesday, on both web and iOS. That kind of speed is where AI builders shine.

The money side is different from most indie stories. FlagStat raised a pre-seed round, which was fully subscribed. A seed round opened in August 2026. So this is a venture-backed company, not a bootstrapped side project. That matters when you compare it with solo apps on this site.

How the first users came

The first big users came through one event. FlagStat was beta-tested at the Florida state flag football championships. It then became the official scorekeeping app of those playoffs.

From there, the claims grow. The Bolt story, published in September 2026, mentions customers in 16+ states, up from 8 earlier that year. The homepage today claims 1,500+ coaches across 30+ states. It also lists FlagStat as an official MaxPreps stat import partner since March 2026. MaxPreps is a widely used US high school sports stats site.

Pricing is freemium. The free tier covers play-by-play stats, box scores, brackets, league schedules and MaxPreps export. Premium adds unlimited cloud game saves, more photo storage and season tools. The homepage does not list a Premium price.

What a non-dev should copy

• Build for rules you know cold. Domain knowledge is the moat AI cannot write for you.
• Win one official event. A single championship put FlagStat in front of every coach who mattered.
• Share one backend across apps. If you use two AI builders, point both at the same database.
• Ship on a fixed weekly cycle. Users forgive gaps when they see requests land within days.
• Put partner badges on the homepage. Official status sells better than any ad copy you write.

If you vibe-code this

Youth sports apps hold rosters, player names and stats for minors. That raises the bar. Check these:

1. Turn on Row Level Security (RLS) for every table. RLS is Supabase's rule layer. Without it, the public "anon" key in your app can read everything. See Supabase RLS for vibe coders.
2. Scope access by role. Coaches, league directors and parents should see different data. Test that changing an ID in a request does not reveal another team. See IDOR in API routes.
3. Keep player photos in a private bucket. Serve them with short-lived signed links. See storage bucket security.
4. Collect the minimum. A stat line needs a jersey number and name, not a birthday or home address.
5. Remember the mobile app ships your keys. Anything in a phone app is public. See React Native secure storage.

An RLS policy so coaches only read players on their own teams:

alter table players enable row level security;

create policy "Coaches read their own team's players"
on players for select
to authenticated
using (
  team_id in (
    select team_id from team_staff
    where user_id = (select auth.uid())
  )
);

Building a sports or youth-data app and want the access rules checked before a season starts? Email me at [email protected] for a security audit.

Key takeaway

FlagStat shows a non-developer can build a real, officially adopted product with Bolt.new. The AI wrote the software. The founder's knowledge of the sport and its gatekeepers won the users.

More case studies: vibe-coded apps making money. Before you launch: the vibe coding security guide.

Sources

• https://bolt.new/blog/flagstat-flag-football
• https://flagstat.games