████████╗ ██████╗ ██████╗ ██████╗ ███████╗██╗ ██╗███████╗ ╚══██╔══╝██╔═══██╗██╔══██╗ ██╔══██╗██╔════╝██║ ██║██╔════╝ ██║ ██║ ██║██████╔╝ ██║ ██║█████╗ ██║ ██║███████╗ ██║ ██║ ██║██╔═══╝ ██║ ██║██╔══╝ ╚██╗ ██╔╝╚════██║ ██║ ╚██████╔╝██║ ██████╔╝███████╗ ╚████╔╝ ███████║ ╚═╝ ╚═════╝ ╚═╝ ╚═════╝ ╚══════╝ ╚═══╝ ╚══════╝

Campus Flow

A campus productivity platform — roommate matching, notes storage, resume builder, PDF chat, and support tickets.

---

Getting Started

Prerequisites

• Node.js v18+
• pnpm or npm

Installation

npm install

Setup Environment

Create a .env.local file in the root with the following:

NEXT_PUBLIC_SUPABASE_URL=your_supabase_url
NEXT_PUBLIC_SUPABASE_ANON_KEY=your_supabase_anon_key
OPENAI_API_KEY=your_openai_api_key

# Optional model defaults for PDF Studio (chat/slides/voice/video)
CHAT_TRANSFORMATION_MODEL=openai/gpt-4o-mini
CHAT_TTS_MODEL=tts-1
CHAT_TTS_VOICE=alloy

Run Development Server

npm run dev

Open http://localhost:3000 in your browser.

Build for Production

npm run build
npm start

---

Database Setup

Run the SQL scripts in order in your Supabase SQL Editor:

1. scripts/setup-db.sql — creates all tables
2. scripts/rls.sql — sets up Row Level Security policies

Resume LaTeX Preview/Download Setup

Resume PDF endpoints:

• POST /api/resume/preview (live preview PDF)
• POST /api/resume/compile (download PDF)

Both compile LaTeX on the server.

Engine behavior

• classic and modern generally use pdflatex.
• compact may require xelatex (auto-detected from template content).
• Binary override env vars:
  • LATEX_BIN_PDFLATEX
  • LATEX_BIN_XELATEX
  • fallback: LATEX_BIN

Required environment variables

Add to .env.local / Vercel / Render as needed:

# Optional local override (fallback for both engines)
LATEX_BIN=/Library/TeX/texbin/pdflatex

# Optional explicit binaries
LATEX_BIN_PDFLATEX=/Library/TeX/texbin/pdflatex
LATEX_BIN_XELATEX=/Library/TeX/texbin/xelatex

# If set, resume compile/preview routes offload compilation to Render
# Example: https://campus-flow-render.onrender.com
LATEX_RENDER_URL=

# Optional full override for remote compile endpoint
# Defaults to {LATEX_RENDER_URL}/api/latex/compile
LATEX_RENDER_COMPILE_URL=

# Shared secret between app and compile endpoint
LATEX_RENDER_SECRET=

Local development requirements

1. Install TeX distribution (MacTeX/TeX Live).
2. Ensure both are available:
   • pdflatex
   • xelatex (important for many compact templates)
3. Run app:
   • npm run dev
4. Test both preview and download from Resume page.

Vercel + Render setup (recommended)

Use hybrid mode:

• Vercel: UI + auth + API routes
• Render: LaTeX compile runtime (/api/latex/compile)

Steps:

1. Deploy same app on Render with TeX installed.
2. Keep /api/latex/compile reachable on Render.
3. Set LATEX_RENDER_SECRET on both Vercel and Render.
4. Set LATEX_RENDER_URL on Vercel to Render base URL.
5. Do not set LATEX_RENDER_URL on Render (prevents proxy loops).

Common preview/download errors

• LaTeX compiler not found...

  • Install TeX binaries or set LATEX_BIN_*.

• LaTeX dependency missing: <file>.sty/.cls

  • Template package/class missing in runtime. Install TeX packages or use Render offload.

• Failed to compile LaTeX

  • Usually template syntax/package/font issue. Check API error detail field.

---

PDF Chat + Studio (Slides, Voice, Video)

/dashboard/chat now has:

• Chat tab: ask questions from selected PDF
• Create tab: generate from selected PDF:
  • Slides (markdown slide deck)
  • Voice (podcast script + mp3)
  • Video Plan (scene-by-scene storyboard markdown)

Requirements

1. PDF must be uploaded and indexed into embeddings for the selected pdf_id.
2. OPENAI_API_KEY must be set for voice generation (TTS).
3. Optional model envs:
   • CHAT_TRANSFORMATION_MODEL (default: openai/gpt-4o-mini)
   • CHAT_TTS_MODEL (default: tts-1)
   • CHAT_TTS_VOICE (default: alloy)

API

• POST /api/chat -> PDF Q&A
• POST /api/chat/generate -> { type: "slides" | "voice" | "video", pdfId, prompt }

Multi-provider model selection (Create tab)

In /dashboard/chat -> Create, you can set a model in provider/model format, e.g.:

• openai/gpt-4o-mini
• anthropic/claude-3-5-sonnet-latest
• google/gemini-2.0-flash
• groq/llama-3.1-70b-versatile
• ollama/qwen2.5

Voice output currently uses OpenAI TTS for mp3 synthesis after script generation.

---

Open Notebook (Embedded in Campus Flow)

Open Notebook can run inside Campus Flow at /dashboard/notebook using rewrite proxies.

Docker (local)

Open Notebook UI should be reachable at:

• http://localhost:8502

Open Notebook API should be reachable at:

• http://localhost:5055

Optional environment overrides

OPEN_NOTEBOOK_URL=http://localhost:8502
OPEN_NOTEBOOK_API_URL=http://localhost:5055

These values are used by:

• health endpoint: GET /api/open-notebook/health
• Next.js rewrites:
  • /open-notebook/:path* -> Open Notebook UI
  • /open-notebook-api/:path* -> Open Notebook API

---

⚠️ License & Usage

Do not copy, reuse, or redistribute any part of this codebase without explicit written permission from the authors.

© 2026 Campus Flow. All rights reserved.