Resona App

Resona is a voice-first dating app that flips the script: hear someone before you see them. Through voice prompts and AI analysis, Resona helps you find someone who truly "sounds like home."

Comment

Inspiration

Resona was inspired by the desire to create a more meaningful and human-centered approach to online dating. Modern dating apps often prioritize visual-first interactions, reducing people to photos and short bios. This leads to shallow connections, ghosting, and dating fatigue. We wanted to change that by focusing on the most human signal of all: voice. How someone speaks — their tone, warmth, and pauses — reveals more about compatibility than any curated photo ever could.

The Boson AI x Eigen AI Hackathon provided the perfect opportunity to bring this vision to life. By leveraging Boson AI’s advanced speech-to-text and vocal behavior analysis tools alongside Eigen AI’s powerful reasoning and content analysis capabilities, we created a system that truly understands users at a deeper level.

What it does

Resona flips the traditional dating app model by prioritizing voice over visuals. Instead of swiping on photos, users answer a series of voice prompts designed to reveal their personality, values, and emotional depth. The app analyzes both what you say and how you say it to create a six-dimension psychological profile.

Resona’s matching algorithm, powered by Boson AI and Eigen AI, pairs users based on vocal and psychological compatibility, helping them find connections that resonate on a deeper level. Once matched, users can continue to build their connection through voice messages and calls, with insights provided by the app’s Vibe Progress engine.

How we built it

Resona is a multi-faceted project that combines cutting-edge technology with thoughtful design:

  • Frontend: The iOS app is built using SwiftUI, providing a seamless and intuitive user experience. The app features a 5-tab interface for onboarding, matching, messaging, and voice interactions.
  • Backend: The backend is powered by FastAPI, handling user profiles, voice uploads, and matchmaking logic. SQLite is used for lightweight data storage during development.
  • Audio Intelligence: The heart of Resona lies in its audio analysis pipeline, powered by Boson AI and Eigen AI:
    • Boson AI: Handles speech-to-text transcription and vocal behavior analysis, extracting insights like warmth, energy, and tonal shifts.
    • Eigen AI: Performs reasoning and content analysis, generating psychological profiles based on the themes, values, and emotional vocabulary in user responses.
  • Matching Algorithm: Our matching system weights vocal and psychological compatibility at 70%, with shared interests at 30%. Confidence-aware scoring ensures that uncertain dimensions are down-weighted rather than guessed.

Challenges we ran into

Building Resona came with its fair share of challenges:

  • Voice analysis: Integrating Boson AI and Eigen AI into a seamless pipeline required careful orchestration to ensure reliable outputs and low latency.
  • Scalability: While our current system works well for a small user base, scaling to thousands of users will require implementing more efficient retrieval methods like FAISS or ScaNN.
  • Privacy concerns: Handling sensitive voice data responsibly was a top priority. We implemented strict data encryption and anonymization protocols to protect user privacy.
  • Balancing simplicity and depth: Designing an app that feels simple and intuitive while delivering deep insights was a constant balancing act.

Accomplishments that we're proud of

We’re incredibly proud of what we’ve achieved with Resona:

  • Boson AI x Eigen AI integration: Successfully combining Boson AI’s vocal behavior analysis with Eigen AI’s reasoning capabilities to create a robust voice-first profiling system.
  • Innovative matching model: By prioritizing voice and psychological compatibility, we’ve created a matching system that goes beyond surface-level attraction.
  • Seamless user experience: The SwiftUI-based frontend provides an intuitive and engaging experience for users.
  • Vibe Progress engine: Our post-match analysis tool provides users with actionable insights to nurture their connections.

What we learned

Throughout the development of Resona, we learned:

  • The power of voice: Voice carries emotional depth and nuance that text and photos cannot capture. We learned how to analyze vocal patterns to uncover personality traits and emotional states.
  • AI integration: Combining Boson AI and Eigen AI taught us how to leverage the strengths of multiple AI systems to create a unified and impactful solution.
  • User-centric design: Creating an app that feels warm and inviting required us to deeply understand user needs and frustrations with existing platforms.
  • AI and ethics: Leveraging AI for voice analysis raised important questions about privacy, consent, and fairness. We worked hard to ensure our algorithms are transparent and respectful of user data.

What's next for Resona App

Resona is just the beginning. We envision a future where voice-first interactions become the norm for building deeper, more authentic connections. Our roadmap includes:

  • Multilingual support: Expanding accessibility to users worldwide.
  • Hybrid acoustic analysis: Combining classical DSP techniques with AI for added robustness.
  • Personalized matching weights: Adapting the algorithm based on user engagement patterns.
  • Outcome validation: Collecting real-world match outcome data to refine and validate our algorithms.
  • Scalability improvements: Implementing sub-linear retrieval methods to handle a growing user base.

Resona is more than just a dating app — it’s a step toward a more human way of connecting in the digital age, made possible by the powerful tools provided by Boson AI and Eigen AI.

Built With

Share this project:

Updates