• 

  (this was transcribed and filled by Whisper) Users can input voice/locket/text about a notable memory

• 

  Interactions are stored in a knowledge graph

• 

  Suggested quest to reach out to loved ones for the week

• 

  Suggested quest to reach out to loved ones for the week

• 

• 

  View your relationship

bond - Relationship Quest Agent

🎯 Inspiration

Work was drowning out my relationships. I built this to actively care for the people who matter—without letting everything else consume me.

Young people already use Locket and BeReal to stay connected. But these are passive. This is the next evolution: AI that helps you actively care about others while you're already sharing moments.

🌟 Impact

Social Health: Combats loneliness by making relationship maintenance effortless

Productivity: Reduces mental load so you can focus on meaningful interactions

For Gen Z: Builds on Locket/BeReal patterns while adding proactive care

💡 What It Does

An AI-powered relationship assistant that:

• Captures memories like Locket (photo/voice/text)
• Tracks relationships in a knowledge graph
• Generates quests telling you WHO to reach out to, WHY, and WHAT to say
• Coaches messages personalized to each relationship

🛠️ How We Built It

• Backend: FastAPI (Python)
• AI: OpenAI GPT-4o (vision, extraction), Whisper (voice)
• Database: Neo4j knowledge graph
• Architecture: Multi-agent AI system (Ingestion, Memory, Relationship Analyst, Quest Generator, Message Coach)

🎨 Key Features

1. Locket-style capture: Photo/voice/text memory input
2. Smart extraction: AI extracts people, events, tasks automatically
3. Relationship tracking: Knowledge graph tracks strength, history, context
4. Proactive quests: "Check in on Alice about her exam next week"
5. Message coaching: Personalized suggestions based on relationship history

🚀 Challenges

• Knowledge graph ontology design
• Balancing AI automation with human review
• Quest prioritization algorithms
• Real-time image/voice processing

🏆 Accomplishments

• Complete end-to-end system in 42 hours
• Sophisticated relationship knowledge graph
• Multi-agent AI architecture
• Intuitive UI for relationship management

🔮 What's Next

• Mobile app
• Email/calendar integrations
• Advanced relationship prediction
• Freemium SaaS model

Track: Productivity & Innovation

Built at DandyHacks 2025 ❤️

Built With

• css
• fastapi
• gemini
• javascript-frameworks:-fastapi
• knowledgegraph
• langchain-core-frontend:-html
• langgraph
• neo4j
• pydantic
• python
• uvicorn
• uvicorn-database:-neo4j-(knowledge-graph)-apis:-openai-(gpt-4o
• whisper
• whisper)-libraries:-pydantic