Studyond — AI Thesis Journey

"From 'I'm starting' to 'I'm submitting'" — Your AI-powered companion for the academic milestone.

---

🌟 The Vision

Studyond is a comprehensive platform designed to streamline the complex process of writing an academic thesis. Built during Start Hack 2026 for the Studyond Challenge, our solution bridges the gap between students, supervisors, and industry partners, transforming a stressful academic milestone into a structured and rewarding journey.

---

🏆 The Challenge: Studyond @ Start Hack 2026

The "Studyond Challenge" at Start Hack 2026 tasked participants with reimagining the academic thesis experience. Our research identified key pain points:

• Fragmentation: Students struggle to find topics across disparate portals.
• Opacity: The registration process is manual and progress is often invisible to supervisors.
• Isolation: A long writing phase with infrequent feedback loops.
• Communication Friction: Students are frequently frustrated with the style, frequency, and consistency of supervisor communication.

Our Solution

We built an end-to-end journey tracker that addresses these through:

• Smart Discovery: Filter topics by Field and Source (Industry/Company vs. University).
• 10-Stage Lifecycle: A visual timeline spanning exploring → application_pending → registered → planning → executing → writing → submitted → defense_prep → graded.
• Proactive AI (Ona): An integrated copilot providing contextual prompts, health checks, and proactive nudges if progress plateaus.
• Unified Supervisor Dashboard: A portal for supervisors to manage approvals, track health scores, and maintain a Dual-Scope Knowledge Base (Global vs. Topic-specific).

---

🛠 Technical Documentation & Design Decisions

🏗 Hybrid Full-Stack Architecture

The platform has evolved into a robust full-stack application:

• Core: React 19 with Vite 6 for the interactive frontend.
• Backend Proxy: A Node.js/Express server (server.js) that handles intelligent API proxying, security headers, and static asset serving.
• AI Intelligence: Powered by Google Gemini 2.5 Flash (@google/genai) for real-time academic support.

---

🧠 Ona AI: The Smart Copilot

Ona is now backed by a live Large Language Model (LLM), enabling:

1. Automated Supervisor Interceptor: A sophisticated logic layer where Ona evaluates student questions in real-time. If an answer exists in the Knowledge Base (formatting, rules, deadlines), Ona provides it immediately, potentially intercepting the message to save supervisor time.
2. Contextual Grounding: Responses are strictly grounded in the supervisor's Knowledge Base to ensure academic rigor.
3. Proactive Health Checks: Automated monitoring of thesis progress to identify and nudge students when they hit plateaus.

---

🛡 Security & Reliability

Designed for high-quality academic data management:

• Helmet.js: Implements essential security headers (CSP, XSS protection).
• Express-Rate-Limit: Protects LLM/API endpoints from abuse and brute-force attempts.
• Dynamic Threading: Maintains state-aware chat histories for seamless student-AI interactions.

---

🎨 Design Guideline (Studyond System)

The application follows a strict, premium design system to ensure a cohesive and high-end academic experience:

• Typography: Uses Avenir Next as the primary typeface for its modern, clean legibility.
• Type Scale: A standardized scale ranging from ds-caption (12px) for metadata to ds-title-xl (36px) for headers.
• Layout Patterns:
  • Adaptive Grids: Responsive 3-column and 4-column layouts for topic discovery and dashboards.
  • Narrow Focus: Content areas restricted to ds-layout-narrow (3xl) to maximize readability during writing and reading phases.
• Color Strategy: Fully implemented using OKLCH color tokens, enabling future-proof dynamic themes and seamless dark mode transitions.
• Glassmorphism Accents: Subtle use of backdrop-blur for high-priority overlays and modals.

---

⚙️ Setting Up

Environment Variables

1. Create a .env file in the root directory (refer to .env.example).
2. Obtain a Google Gemini API Key from Google AI Studio.
3. Add your key:

   GEMINI_API_KEY=your_actual_key_here

Local Development

# Install dependencies
npm install

# Run Vite dev server (frontend only with proxy)
npm run dev

# Run Production-ready full-stack server
npm start

---

🚀 Deployment Guide (Railway)

We've chosen Railway for its superior DX and reliability.

Step-by-Step Deployment

1. Create Project: + New Project > Deploy from GitHub repo.
2. Configure Environment: Add GEMINI_API_KEY to the project variables in Railway.
3. Configure Build Settings:
   • Start Command: npm start (Railway will automatically detect server.js)
   • Build Command: npm run build
   • Install Command: npm install
4. Network: Ensure the port is set to 8080 (or let Railway use the default PORT variable).

---

👥 The Team

Created at Start Hack 2026 for the Studyond Challenge.

Special thanks to the Studyond mentors for the challenge insights.

---

Note

This project is a prototype built for the Start Hack 2026 hackathon.