Papegaai.ai

Turn — Universal Game Engine

AI Integration into Processes

Challenge

Building a new board or card game requires implementing the same infrastructure repeatedly: legal move generation, game state management, AI player types, and statistical analysis of rule changes. Without a shared engine, each prototype is a greenfield project and designers cannot get fast feedback on whether a rule change actually improves the game.

What we built

We built a generic engine where each game is expressed as a rule file: legal moves per state, win and draw conditions, scoring. Players are pluggable — a single match can mix random, heuristic, AI-policy, and GTO-solver bots. Built-in metrics (luck factor, balance, game length) run automatically across thousands of self-play games whenever a rule changes. Adding a new game means writing one rule file and getting analysis for free.

Result

The engine underlies Flipping Poker and a set of test games. Each new game variant gets automated metrics within minutes of a rule change — designers see the effect of their tweak before committing to it. The architecture makes expanding the game portfolio a matter of writing rule files, not rebuilding infrastructure.

Tech used

delivered within 7 days

Free Prototype