Whodunit AI: AI-Powered Interactive Murder Mystery Generator

Whodunit AI is a cinematic AI murder mystery game where players investigate clues, interrogate suspects, and solve uniquely generated cases in real time.

Comment

Inspiration

Most mystery games eventually become predictable because the stories are fixed. We wanted to create a detective experience where every mystery feels fresh, replayable, and emotionally immersive using AI-generated storytelling and gameplay.

What it does

Whodunit AI is a cinematic AI-powered murder mystery game where players:

  • Generate unique mystery cases
  • Investigate crime scenes
  • Reveal clues progressively
  • Interrogate AI-driven suspects
  • Reconstruct timelines
  • Make final accusations

Every mystery is dynamically generated, making each playthrough different.

How we built it

We used MeDo to rapidly prototype and build the application experience. The project was designed using:

  • TypeScript
  • Tailwind CSS
  • Framer Motion
  • Zustand for state management
  • Supabase for Backend Storage
  • AI-generated mystery and interrogation flows

MeDo helped accelerate UI generation, gameplay structuring, cinematic transitions, and interaction design.

Challenges we ran into

One of the biggest challenges was balancing AI creativity with gameplay consistency. We needed suspects, clues, motives, and timelines to remain logically connected while still feeling dynamic and unpredictable.

Designing suspenseful interactions and maintaining immersion across mobile and desktop layouts was also challenging during rapid prototyping.

Accomplishments that we're proud of

We are especially proud of:

  • The cinematic investigation experience
  • Progressive clue reveal mechanics
  • AI-powered suspect interrogations
  • The dramatic reveal sequence
  • Building a polished playable prototype rapidly using MeDo

The experience feels more like participating in a live mystery than reading a static story.

What we learned

We learned that AI can go beyond content generation and become a core gameplay mechanic. We also learned how powerful AI-native development workflows can be for rapidly validating interactive product ideas.

What's next for Whodunit AI

Next, we want to expand Whodunit AI with:

  • Multiplayer mystery sessions
  • Persistent player profiles
  • Community-generated mystery themes
  • Voice-based interrogations
  • AI-generated visual evidence
  • Daily challenge mysteries
  • Advanced detective progression systems

Our long-term vision is to create a fully AI-driven interactive detective universe with endless replayability.

Built With

  • medo
  • supabase
  • zustand
Share this project:

Updates