IRL Journaly

Welcome to IRL Journaly! This is my graduate project for solirius reply that uses Ruby on Rails, 3D Gaussian Splatting (ml-sharp) and AI text processing (Groq).

Follow the instructions below to get your local development environment set up.

Prerequisites

• Ruby (v3.1.2 or higher recommended)
• PostgreSQL
• Python 3.11+ (for ml-sharp)

---

1. Download / Clone the Repository

First, clone this repository to your local machine and navigate into the project folder:

cd Desktop
git clone https://github.com/jes4l/irljournaly.git
cd irljournaly

2. Set up ML-Sharp (3D Gaussian Splatting)

IRL Journaly relies on Apple's ml-sharp to convert 2D images into 3D Gaussian Splats. You can check out their official repository here for more details and documentation: Apple ML-Sharp on GitHub.

You need to download it and set up a Python virtual environment to run it locally alongside the Rails app.

1. Clone the ml-sharp repository somewhere on your computer (e.g., your Desktop):

   cd Desktop
   git clone https://github.com/apple/ml-sharp.git
   cd ml-sharp

2. Create the input and output folders (useful for testing the model manually):

   mkdir input_images
   mkdir output_gaussians

3. Create and activate a Python virtual environment (Make sure you name it .venv!): If you are using conda:

   conda create -n sharp_env python=3.11
   conda activate sharp_env
   # Or using standard python venv:
   python3 -m venv .venv
   source .venv/bin/activate

4. Install the dependencies required by ml-sharp:

   pip install -e .

3. Update the ML-Sharp Path in Rails

You need to tell the Rails application where you downloaded ml-sharp.

1. Open irljournaly/app/jobs/generate_splat_job.rb in your code editor.
2. Find the following line (around line 26):

   ml_sharp_path = "/Users/jesal.vadgama/Desktop/ml-sharp"

3. Change this path to the exact absolute path where you cloned ml-sharp in Step 2.

4. Get Your API Keys & App Passwords

To use the AI transcription and email features, you need to generate a couple of keys.

Groq API Key (For AI Transcripts)

1. Go to the GroqCloud Console.
2. Create an account or log in.
3. Click "Create API Key", give it a name, and copy the generated key.

Google App Password (For Email Sending)

Because standard Gmail passwords don't work for SMTP servers, you need an App Password.

1. Go to your Google Account Security Settings.
2. Ensure 2-Step Verification is turned on.
3. Search for "App Passwords" in the settings search bar (or go to 2-Step Verification -> App Passwords).
4. Create a new App Password and name it your website name (e.g., jesal.wiki).
5. Copy the 16-character password generated.

5. Set up your .env File

In the root folder of your irljournaly Rails project, create a new file called .env and add the keys you just generated:

# .env

GROQ_API_KEY=your_groq_api_key_here
GMAIL_USERNAME=your_email@gmail.com
GMAIL_APP_PASSWORD=your_16_character_app_password_here

6. Install Dependencies and Run

Now that everything is configured, install the Ruby dependencies and set up your database.

1. Install the gems:

   bundle install

2. Prepare the database:

   bin/rails db:prepare

3. Start the application: IRL Journaly uses a custom host script to boot up the Rails Server, Solid Queue (for background jobs), and Cloudflare Tunnels simultaneously.

   ./bin/host

   (Note: If you are not using Cloudflare Tunnels, you can just start the server normally with bin/rails s and run bin/jobs in a separate terminal).

Visit http://localhost:3000 (or your tunnel URL) in your browser, and you're good to go!

7. Next Steps (Planned Features)

Here are some of the upcoming features planned to make IRL Journaly even better:

• Daily Journaling Reminders: A notification system (via email or web push) to gently remind you if you haven't logged an entry for the day.
• Interactive Doodle Canvas: A built-in drawing tool that lets you sketch and doodle directly onto your journal pages alongside your text and 3D models.
• Strict Splat Validation: Update the Gaussian Splatting background job to strictly accept only valid image files, preventing accidental processing of unsupported formats.
• Media Support: Expand the journal canvas to accept and seamlessly play all file types, including audio files, videos, and GIFs.
• Custom Audio Override: Add an option to replace the AI-generated Text-to-Speech (TTS) transcript audio with your own uploaded voice memos and recordings.