Grandma's Boy

Grandma’s Boy is a voice-based AI application designed to preserve family history by capturing and organizing a grandmother’s stories directly from everyday phone calls. The app automatically calls grandma, engages her in natural, meaningful conversations, and uses advanced AI to structure her anecdotes into cohesive narratives that family members can revisit anytime.

The app addresses a common family challenge: priceless stories from older generations are often lost or fragmented over time. Traditional solutions like written journals or manual recordings place a burden on family members and often result in incomplete, disorganized recollections. Grandma’s Boy eliminates this friction by automating the entire process.

Key Features:
• Automatic Story Capture: The app calls grandma on a scheduled basis, guiding her through conversations and extracting important memories and life stories.
• AI-Powered Story Structuring: Using OpenAI’s o1-mini model, the app analyzes transcribed conversations to identify, structure, and compile stories into well-organized categories (e.g., “Childhood Adventures,” “Family Legends”).
• Voice Playback: Family members can listen to stories in grandma’s authentic voice, providing a direct, emotional connection to her memories.
• Cross-Call Memory: The app retains context across conversations, allowing grandma to revisit and build upon past stories naturally.
Innovation & Creativity: While voice-based AI companions are not new, Grandma’s Boy uniquely combines real-time conversational AI with automated memory structuring. Rather than acting as a generic voice assistant, it becomes a personalized family historian that captures, refines, and preserves authentic narratives effortlessly.

Real-World Impact: Family stories are central to intergenerational connection. By making grandma’s life stories easily accessible and shareable, the app strengthens family bonds while preserving invaluable memories for future generations.
Theme Alignment: The project demonstrates cutting-edge AI by leveraging OpenAI’s real-time voice API and GPT-4 reasoning capabilities to transform unstructured conversations into cohesive, meaningful narratives.

Technologies Used:
• OpenAI Realtime API: Voice interaction with natural, fluid conversation.
• OpenAI GPT-4o1: Memory-aware storytelling and narrative extraction.
• AssemblyAI: Speech-to-text transcription for recorded calls.
• Twilio: Telephony infrastructure for automatic, interactive phone calls.
• Python (Custom Code): Sentence extraction, cleanup, and story organization across calls.
• Pipecat: Voice interaction routing.
• o1-mini: Optimization for efficient conversation flow.
• Opik for monitoring the AI requests