JournIA

AI-assisted journaling with emotion analysis, visuals, and manual mood tracking.

---

Overview

JournIA is a full-stack app that helps you journal and visualize your mood data.

• Journal: free-form entries with AI integration.
• Graphs: interactive dashboards for the analyses.
• Manual Metrics: add your own activities, spider quiz ratings, and Plutchik emotions.
• Past Entries: review previous conversations and their analysis.
• Notes: a simple notes area.
• Settings: configure your preferences.

What you’ll see when you run it

1. Home → enter age and gender (used by analysis).
2. Dashboard → the hub. From here you can jump to:
   • Journal (write a new entry -> run AI analysis),
   • Graphs (visualize analyses; can filter by day/week/month or by entry),
   • Past Entries (open a specific conversation to see its metrics),
   • Add Metrics (manually add mood/quiz/emotions),
   • Notes and Settings.

---

Requirements

• Python: 3.11 (3.11.9 recommended if using GPU llama-cpp-python build)
• Node.js: ≥ 18 LTS recommended
• npm: ≥ 9
• (Optional) CUDA: recent version if you plan to run llama-cpp-python with GPU
• (Apple Silicon) Homebrew with cmake and pkg-config

---

Installation

Create & activate a virtual environment from inside JournAI/ (recommended):

# macOS / Linux
python3.11 -m venv .venv
source .venv/bin/activate

# Windows
py -3.11 -m venv venv
.\venv\Scripts\activate

Run following commands to install dependencies:

Backend:

• pip install -r requirements.txt
• $env:CMAKE_ARGS="-DGGML_CUDA=on"; pip install --force-reinstall --no-cache-dir llama-cpp-python —> to use GPU (standard llama-cpp-python is for cpu only. For this one we need python version 3.11.9 and a recent version of CUDA)
• brew install cmake —> for Apple M1 hardware
• brew install pkg-config —> for Apple M1 hardware

Frontend:

• npm install
• npm install -g @angular/cli

Running the app

• Inside backend run: uvicorn main:app --reload to start the server
• Inside frontend run: ng serve to start the web cli