Skip to content

bhaktofmahakal/ai-counsellor-hf

Repository files navigation

🎓 AI Counsellor — Your Strategic Study Abroad Mentor

Vercel Deployment Next.js Groq

AI Counsellor is a guided, stage-based platform designed to turn student confusion into clarity. Unlike generic chatbots, this is an Execution-led Decision Engine that guides students through a strict 4-stage roadmap: from profile building to university locking and application preparation.


🚀 Live

Experience the future of education counselling: https://ai-counsellor-hf.vercel.app/


✨ Core Features

🧠 1. Agentic AI Counsellor (Call Mode)

Powered by Groq (Llama 3.3 70B) for ultra-low latency (<500ms).

  • Interactive Voice: Speak naturally with your mentor using "Call Mode".
  • Agentic Actions: The AI doesn't just talk; it creates tasks, shortlists universities, and drafts documents directly in the database using custom action tags.

🔍 2. RAG-Powered Discovery

Built using Upstash Vector DB for semantic search.

  • Contextual Matching: Finds universities based on the "intent" of your search, not just keywords.
  • Match Scoring: Every university is ranked as Dream, Target, or Safe based on a custom algorithm comparing student GPA, budget, and goals.

🛡️ 3. Decision Discipline (University Locking)

  • Commitment Step: Users must "Lock" a university to proceed to the application phase.
  • Stage-Gate Logic: The platform restricts access to specialized tools (like the SOP drafter) until a commitment is made, ensuring focus and momentum.

📊 4. Mission Control Dashboard

  • 4-Stage Roadmap: Visual tracking of your journey (Profile → Discovery → Finalizing → Application).
  • Profile Strength: Real-time calculation of your "Precision Score" to identify gaps in your academic profile.

📄 5. AI Document Vault

  • Auto-Generated SOPs: AI drafts your Statement of Purpose based on your unique profile.
  • Export to PDF: Refine your documents in a built-in workspace and export professional PDFs.

🛠️ Technical Architecture

Layer Technology
Frontend Next.js 15 (App Router), Tailwind CSS, Framer Motion
Backend Next.js API Routes, Prisma ORM
Database PostgreSQL
AI/LLM Groq (Llama 3.3 70B), OpenAI Whisper (for STT Fallback)
Vector DB Upstash Vector (RAG Implementation)
Auth NextAuth.js (Google & Credentials)
State Management Zustand

📸 UI Aesthetics

  • WebGL Shaders: LiquidChrome effects for a premium "SaaS" feel.
  • Bento Grids: Modern, clean organization of data and stats.
  • Micro-interactions: High-fidelity animations using Framer Motion.

⚡ Quick Start (Local)

  1. Clone the repo:

    git clone <repo-url>
    cd ai-counsellor
  2. Install dependencies:

    npm install
  3. Set up environment variables (.env):

    DATABASE_URL=
    GROQ_API_KEY=
    UPSTASH_VECTOR_REST_URL=
    UPSTASH_VECTOR_REST_TOKEN=
    NEXTAUTH_SECRET=
  4. Initialize Database:

    npx prisma generate
    npx prisma db push
  5. Run the dev server:

    npm run dev

🎯 Our Vision

To democratize high-end education consulting. We believe every student deserves a mentor who understands their data, respects their goals, and drives them toward execution.

About

AI Counsellor is a guided, stage-based platform designed to turn student confusion into clarity.

Resources

Stars

1 star

Watchers

0 watching

Forks

Packages

 
 
 

Contributors