Intimity

• 

  Logo

• 

  Intimity

Inspiration

We often talk to our partners every day, but how often do we really understand what we’re communicating? Intimity was born from that gap, a desire to uncover the emotional patterns hidden in our everyday conversations. Inspired by sentiment analysis tools and relationship psychology, I wanted to build something that helps people reconnect through understanding, not just messaging.

What it does

Intimity is a relationship sentiment analyzer that turns exported chat history into rich emotional insight. Users can upload their conversations, and Intimity generates sentiment timelines, emotional heatmaps, and highlights around love, conflict, and communication patterns. It also includes a feature that uses AI to help craft personalized messages, from apologies to words of affirmation based on your real relationship tone. It’s like emotional intelligence, visualized.

How we built it

I built Intimity’s frontend using Bolt.new, with React, Tailwind, and Framer Motion for animations. I used Supabase to store anonymized emotional metadata, never touching raw message content after processing. For AI, I integrated Gemini's API to run sentiment scoring and generate smart, emotionally tuned message suggestions. The core dashboard brings together emotional metrics, interactive charts, and soft UI design that matches the sensitivity of the use case.

Challenges we ran into

Parsing different chat formats (like WhatsApp and Telegram exports) was trickier than expected, especially maintaining time-based coherence. Designing for privacy was another core challenge. I had to find a way to analyze deeply personal data without storing any personally identifiable information. Finally, translating raw emotion into elegant UX required constant testing, I didn’t want the product to feel too clinical, or too vague.

Accomplishments that we're proud of

I’m proud that Intimity delivers real emotional insight in a form people instantly understand. From timelines to love expressions, everything feels intentional and personal. I’m also proud of the identity I designed, a visual brand that balances data with empathy. Perhaps most of all, I’ve already seen people respond with curiosity and connection. That’s what this project is about.

What we learned

This project reminded me that building with AI isn’t about complexity, it’s about clarity. I learned how to create emotionally aware interfaces that feel both smart and human. I gained practical skills in data visualization, prompt tuning, and privacy-first architecture. And most importantly, I saw how powerful AI can be when it helps people understand themselves and each other better.

What's next for Intimity

Next, I plan to launch a public beta and gather real feedback from couples. I’ll continue to expand the insights with features like relationship forecasting, communication health scores, and expert-informed feedback. I’m also exploring partnerships with relationship coaches and therapists to turn this into a deeper emotional wellbeing tool. And of course, I’ll be building a mobile version to make these insights even more accessible, because emotional understanding should be just a tap away.

Built With

• bolt.new
• elevenlabs
• framer-motion
• gemini
• lucide
• next.js
• openai-api-(gpt-4)
• react
• recharts
• supabase
• tailwind-css
• typescript
• vercel
• zod