Stoix

Landing Page
Red Pill User Input
Red Pill Results
Blue Pill User Input
Blue Pill Results

Inspiration

Most of our daily routines are not intentional—they’re automatic. As students, we noticed two extremes: when work piles up, people feel overwhelmed and don’t know where to start; when free time appears, it often disappears into passive habits like scrolling or watching content. Existing tools either over-structure your day or leave you to figure everything out yourself.

We were inspired to design something that sits in between—something that helps people actively choose how to spend their time. The Red Pill / Blue Pill metaphor became a natural way to frame this idea: structure vs exploration, discipline vs spontaneity.

What it does

Stoix is a system that helps users make better decisions about how they spend their time.

🔴 Red Pill (Execution Mode):
Breaks down large tasks into smaller steps and schedules them realistically, reducing overwhelm.
🔵 Blue Pill (Exploration Mode):
Suggests real-world activities (“side quests”) based on time, location, budget, and interests—helping users use free time more meaningfully.
🧠 Skill Mode:
Recommends skills users can start immediately based on their available time, with step-by-step session plans and growth paths.

Instead of just listing options, Stoix filters and structures them so everything suggested is actually doable within the user’s time constraints.

How we built it

We built Stoix as a full-stack web application:

Frontend:
- Next.js + React + Tailwind CSS
- Cyberpunk-inspired UI with a dark, minimal, system-like interface
- Interactive landing page centered around choosing between Red and Blue Pill
Backend:
- API routes (/api/red-pill, /api/blue-pill, /api/skill-path)
- Aggregates data from:
- Google Places (cafes, gyms, etc.)
- Event APIs (Ticketmaster/Eventbrite)
- Curated datasets for niche and free activities
AI Layer:
- Used AI models to:
- break down tasks
- rank and filter activities
- generate skill roadmaps
- Enforced structured JSON outputs to keep responses consistent and usable

Challenges we ran into

Time feasibility logic:
Ensuring every suggestion fit within the user’s available time required accounting for travel time, activity duration, and constraints.
API inconsistencies:
Different APIs returned data in different formats, so we had to normalize everything into a single schema.
UI complexity:
Designing a cyberpunk-style interface that still felt clean and usable was challenging—we had to avoid overloading the user visually.
AI reliability:
Getting consistent, structured outputs from models required strict prompts and formatting rules.

Accomplishments that we're proud of

Building a fully working, interactive prototype that connects UI, backend, and AI
Creating a system that feels like a decision engine, not just another productivity tool
Designing a cohesive visual identity that reinforces the concept (Matrix-inspired choice system)
Successfully combining:
- planning
- exploration
- skill-building
  into one unified experience