Build Story · Party game · EN / BG
Quiz Night: run to your answer
The party quiz where your phone is the controller and the TV is the stage.
In one look
The problem
Party quizzes either make everyone install an app or crowd around one screen. Quizzes at work feel like homework.
What we built
Open it on a TV or laptop, friends scan the QR, and pixel avatars run to answer zones A–D. No install, no ads, no third-party analytics.
Where it is now
Live and free in English and Bulgarian: solo, a free 1v1 Quick Duel, party rooms for up to 8 and a Daily Challenge. A native iPhone version is in testing.
What this would have cost in 2024
About $53k and 2–3 months
2024 team: 4 people: game backend, frontend, iOS, pixel artist with QA. Built here by one founder working with AI in 24 days.
| Piece | 2024 hours |
|---|---|
| Multiplayer engine, QR lobby, server clock, anti-cheat | ~288 |
| Native iPhone app + rules parity | ~270 |
| 23 maps, characters, sound | ~207 |
| Quick Duel, ghosts, perks | ~147 |
| Teams branding, accounts, Daily | ~146 |
| Question bank EN/BG + audit | ~143 |
| QA, load tests, security, store prep | ~113 |
| AI quizzes, document quizzes, AI worlds | ~107 |
| Total (PERT) | ~1,421 |
Method: PERT (O+4M+P)/6 per piece; low = O × $25, likely = PERT × $37.5, high = P × $50 (Accelerance 2024 Eastern-Europe band). These are estimates of what a team would have needed in 2024, not invoices.
How it was built
From first commit to current build in 24 days
- First commit
The idea: the TV is the stage and phones are the controllers.
- The whole game in one pass
QR lobby, phone controller, server-driven game clock, answers shuffled per play and kept on the server until the reveal, AI and document quizzes, team branding, chiptune sound.
engine load-tested at 50 players (rooms hold 8) · 428/428 answers accepted - Daily Challenge
The same 8 questions for everyone, new at midnight, with a streak and an emoji grid to share.
- AI worlds, Quick Duel and a question audit
"Create a world" generates a themed map with a host character and its questions. Free 1v1 duels play a clearly labelled ghost when no one is free. All 680 built-in questions were audited.
0 wrong answer keys found - Perks
Freeze Ray, Shield, 50/50 and Double Down, enforced on the server. A rule we kept: perks never turn a wrong answer right.
22 unit tests · scoring without perks unchanged - Full player-side QA, more questions
Five fix packs from a full QA round, 733 new questions in both languages, and a Daily that doesn't repeat within 42 days.
- A native iPhone app
Solo and Daily rebuilt natively, sharing the web game's rules. A parity check proves both play by the same rules.
136/136 rule cases match - Security pass
AI question generation got spending caps that fail closed.
- Duels made fair under load
Answers sent in the first instant are no longer lost, and the duel state call got almost 3× faster.
640/640 answers accepted · 2.2 s → 0.83 s
Who did what
The founder decided
- Rooms cap at 8, so every phone stays readable on the TV.
- Duel opponents who aren't human are labelled as ghosts.
- Perks can't make a wrong answer right.
- No ads and no third-party analytics.
AI built and checked
- The web app, built in Lovable from the founder's briefs.
- Claude Code: perks, the realtime rework, the iPhone app, the rules parity check, load tests and QA rounds.
- The question bank, with an AI audit and human-approved fixes.
Stack
Want something like this for your team?
From one call to something live. The same method is in the Claude Operator Pack.
From the founder
You might also like
Shares TanStack Start · SwiftUI
Shares TanStack Start · SwiftUI
Shares TanStack Start · SwiftUI