-
-
Landing Page/Main Page
-
User Dashboard/Memories Page (Users can access their saved memories and track their progress and milestones)
-
Record Page (Here users can record their voice and use the AI analysis to transcribe and sentiment analysis and save or cancel memories)
-
Insights Page (This page provides an sleek, modern and informative interface to view and trach their memories and overview in one go)
-
Profile Page (An interface for users to edit their information and manage their subscription plan)
-
Settings Page (Here users can switch between different themes and manage their preferences)
-
Subscription/payment Interface (Provides a pop up interface to upgrading the subscription plan)
Inspiration
The inspiration for VoiceVault came from a deeply personal place - watching elderly family members struggle with memory loss while simultaneously realizing how much of our own precious memories we lose to time. We're living in an age where we capture thousands of photos but rarely revisit them, where meaningful conversations fade into forgotten fragments. We envisioned a world where technology could serve as a bridge between our fleeting human experiences and permanent digital preservation. The idea of talking to an AI companion that truly understands and remembers our stories felt like the missing piece in how we interact with our personal history.
What it does
VoiceVault is a revolutionary personal memory assistant that transforms how we capture, organize, and retrieve life experiences through natural voice conversations. Users simply talk to their personalized AI avatar companion, sharing memories, stories, and moments as if speaking to a trusted friend. Core Features:
Voice-First Memory Capture: Natural conversations like "Hey VoiceVault, remember that amazing sunset we saw in Bali with the street food vendor who spoke perfect English?" AI Avatar Companion: Personalized video avatars that respond with emotional intelligence and visual memory playback Blockchain Privacy: Encrypted memory ownership on Algorand with selective sharing capabilities Smart Retrieval: Contextual memory search through conversational queries Premium Insights: AI-powered patterns, anniversaries, and emotional intelligence about your life story
How we built it
Our tech stack strategically integrates multiple challenge sponsors to create a cohesive, scalable solution: Frontend (Bolt.new): ->application with a responsive voice recording interface ->Real-time chat-style memory browser with search capabilities ->Tavus AI avatar integration for personalized video interactions ->RevenueCat subscription management for premium tiers
Backend Infrastructure (Supabase): ->PostgreSQL database for memory metadata and user management ->Real-time subscriptions for instant memory sync ->Row Level Security for privacy protection ->Edge functions for AI processing workflows
AI Integration: ->ElevenLabs: Voice synthesis for natural avatar responses and speech-to-text for memory capture ->Tavus: Real-time AI video avatar generation with emotional expressions ->Custom NLP pipeline for memory entity extraction and semantic search -> Used API from GORQ cloud for AI analysis
Blockchain Layer (Algorand): ->Smart contracts for memory ownership and inheritance ->Encrypted storage with user-controlled access keys ->Micropayment system for premium feature unlocks ->Privacy-preserving selective sharing mechanisms
Deployment (Netlify): ->Continuous deployment pipeline with edge optimization ->Custom domain integration for professional presence
Challenges we ran into
Technical Integration Complexity: Synchronizing real-time voice processing, AI avatar generation, and blockchain transactions while maintaining a smooth UX proved challenging. We solved this with a queue-based architecture that processes heavy operations asynchronously. Privacy vs. Functionality Balance: Implementing true blockchain privacy while enabling semantic search required innovative cryptographic approaches. We developed a system where metadata remains searchable while actual memory content stays encrypted.
Voice Recognition Accuracy: Handling emotional speech, background noise, and varied accents required extensive ElevenLabs API optimization and custom preprocessing pipelines. Cross-Platform Avatar Consistency: Ensuring Tavus avatars maintained personality and visual consistency across different devices and interaction contexts required careful prompt engineering and state management.
Scalable Subscription Logic: Implementing RevenueCat's subscription tiers while maintaining blockchain ownership principles required custom middleware to bridge Web2 payments with Web3 ownership.
Accomplishments that we're proud of
Seamless Multi-Platform Integration: Successfully integrated 6 different sponsor technologies into a cohesive, production-ready application that feels native rather than stitched together. Revolutionary UX: Created the first truly conversational memory interface where users can retrieve specific memories through natural language without any learning curve. Blockchain Privacy Innovation: Developed a novel approach to privacy-preserving memory storage that gives users complete ownership while enabling smart AI insights. Emotional AI Breakthrough: Our Tavus avatars don't just respond - they reflect the emotional significance of memories, creating genuine emotional connections with users. Business Model Validation: Built a clear path from MVP to millions of users with proven monetization through RevenueCat and scalable infrastructure via Supabase.
What we learned
Voice-First Changes Everything: Moving beyond traditional text interfaces to voice-first interactions revealed how much more natural and emotional memory sharing becomes when it mimics human conversation.
Privacy Isn't Binary: Users want control over their data without sacrificing functionality. Our blockchain approach proved that you can have both complete privacy and intelligent AI insights.
AI Avatars Need Personality: Technical capability isn't enough - AI avatars need consistent personality, emotional intelligence, and memory of past interactions to feel authentic.
Monetization Requires Value: Users will pay for premium memory features, but only when they enhance the emotional and practical value of their personal stories.
Integration Amplifies Innovation: Each sponsor technology became exponentially more powerful when thoughtfully integrated with others, creating emergent capabilities none could achieve alone.
What's next for VoiceVault
Immediate Roadmap (3-6 months):
Advanced family sharing features with multi-generational memory trees Integration with photo/video libraries for automatic memory enhancement Therapy and eldercare partnerships for memory assistance applications Mobile app development for seamless cross-device experience
Long-term Vision (6-18 months):
AI-powered memory coaching for improved recall and emotional processing Marketplace for sharing anonymous memory insights for research Corporate knowledge management applications for team memory preservation Integration with AR/VR for immersive memory re-experiencing
Scaling Strategy: Leverage Supabase's global infrastructure for worldwide expansion Develop an API ecosystem for third-party memory applications Create white-label solutions for healthcare and enterprise clients Build community features for shared cultural and historical memory preservation
VoiceVault represents the future of human-AI collaboration in preserving what matters most - our stories, connections, and the memories that make us who we are. We're not just building an app; we're creating a new category of deeply personal, emotionally intelligent technology that grows with users throughout their lives.
Built With
- bolt.new
- groq
- postgresql
- supabase
- typescript
- vite


Log in or sign up for Devpost to join the conversation.