Product Lab Showcase
Same
Social product for anonymous emotional connection and real feelings.
- Tech Stack
- Next.js 16 React 19 Tailwind CSS v4 HeroUI Framer Motion Drizzle PostgreSQL Clerk Pusher
- Project Type
- Personal Product Lab
Product Concept
How it works & why I built it
An anonymous peer-to-peer mental health support application designed to create genuine connection without the pressure of traditional social media.
My Role & Engineering Scope
Product concept, interaction design, OLED UI direction, full-stack implementation, PostgreSQL schema design, realtime chat architecture, gamification, and bilingual i18n.
Problem
Public social products often make vulnerable posts performative or unsafe through comments, follower counts, and identity pressure.
Concept
Same removes the social scoreboard and centers the product around anonymous sharing, mutual consent, and small supportive actions.
Built
- Anonymous posting and feed interactions.
- Double-opt-in matching and private chat flows.
- Mood check-ins, karma, streaks, and bilingual interface support.
Product Highlights
- Comment-free feed focused on empathy and support.
- Swipe-to-match flow that creates double-opt-in private chats.
- Real-time private messaging with typing indicators and presence.
Key Features
Comment-Free Feed
Users can share anonymously while interactions stay limited to Same and Hug reactions.
This reduces trolling, debate, and unwanted advice on sensitive emotional posts.
Swipe-to-Match Connections
Private conversations begin only after a mutual match request.
The double opt-in model protects boundaries before direct messaging starts.
Daily Vibe Check
A dedicated mood-tracking flow with streak behavior and emotional self-observation.
It gives the product daily utility beyond the social feed.
SOS Crisis Mode
Quick-access crisis support content for moments that need immediate help.
A mental health product needs an explicit safety path, not only engagement loops.
UX Decisions
- Used anonymous abstract avatars instead of real photos to reduce ego-driven posting.
- Removed public comments so vulnerable posts are not exposed to debate or backlash.
- Built an OLED-style dark interface to make the product feel private, calm, and secure.
- Structured posting around struggle tags and mood rather than status updates.
Engineering Decisions
- Separated database logic from UI through a clear data access layer.
- Used Pusher private and presence channels for typing indicators and online status.
- Validated payload boundaries with Zod and added Arcjet protection against automated abuse.
- Kept local database iteration fast with PGlite while targeting PostgreSQL in production.
Challenges
- Designing private engagement without opening harassment paths.
- Keeping realtime presence and typing state responsive without excessive database writes.
- Modeling reactions, chats, karma logs, and connection requests in a relational schema.
Outcome
Built a polished social product lab that demonstrates protective UX strategy, real-time product behavior, and robust full-stack execution.