Emome

Inspiration

Most mental-health tools track your mood, but not what (or who) caused it. Real life is made of interactions: a supportive friend, a draining colleague, a confusing situationship. Emome was inspired by the idea that clarity comes from patterns—if you can measure your relationship dynamics over time, you can make better decisions (and stop gaslighting yourself with “maybe it’s nothing”).

What it does

Emome is a privacy-first Emotional Relationship Tracker. In seconds, you log an interaction (who it involved, intensity, duration, and a short note). Emome converts those micro-moments into:

• Relationship Pulse: a time-weighted emotional index per person
• Trends: who consistently nourishes you vs. drains you
• Weekly summaries: what changed and why (without medical claims)
• Reflection prompts: quick follow-ups that make the data cleaner (“energizing or draining?”, “boundary crossed?”)

How we built it

• Frontend: a lightweight web app for fast interaction logging + clean visualizations
• Core logic: a scoring model that weights intensity × duration and decays over time to reflect recency
• AI layer: generates short, non-judgmental summaries and pattern insights from your notes
• Privacy mindset: minimal data collection, user-controlled logs, designed for “no cringe, no oversharing”

Challenges we ran into

• Avoiding “therapy claims” while still being genuinely helpful
• Signal vs. noise: designing a score that doesn’t overreact to one bad day
• UX friction: making logging feel instant (not like journaling homework)
• Insight quality: keeping AI outputs grounded, consistent, and respectful

Accomplishments that we're proud of

• A working end-to-end product that turns messy emotions into readable signals
• A simple but powerful Relationship Pulse model that users can understand
• Clean UI/UX focused on speed: log in seconds, learn in minutes
• Summaries that feel supportive without pretending to be a clinician

What we learned

• Users don’t want “more feelings.” They want clarity and pattern recognition.
• The best insights come from tiny structured prompts, not long diaries.
• Trust is everything: privacy + tone + transparency beats fancy features.

What's next for Emome

• Multi-person dashboard + comparison views (family, friends, coworkers)
• Smarter trend alerts (“this relationship has been net-draining for 3 weeks”)
• Voice-first logging (speak an event, get it structured automatically)
• Export/shareable reports for personal reflection (still privacy-first)

Built With

• localstorage
• next.js
• openai
• react
• tailwind-css
• typescript
• vercel