Installation

How to install SENTINEL\u00B7IEP on Windows, macOS, or Linux.

Requirements

Component	Minimum	Recommended
OS	Windows 10, macOS 12, Ubuntu 22.04	Windows 11, macOS 14+, Ubuntu 24.04
RAM	8 GB	16 GB +
Disk	5 GB free	20 GB free (for larger AI models)
Python	3.10	3.12 or 3.14
GPU	Not required	NVIDIA RTX 3060 or better (8GB+ VRAM)

Quick install (Windows)

1 Install Python 3.12 or newer from python.org. Check "Add Python to PATH" during installation.

2 Open Command Prompt and verify Python is available:

python --version

You should see Python 3.12.x or newer.

3 Install the required Python packages:

pip install faster-whisper sounddevice numpy psutil pillow

This takes about 2 minutes. Faster-whisper is the speech-to-text engine; the rest are supporting libraries.

4 Download SENTINEL\u00B7IEP. [PLACEHOLDER \u2014 once we have a public release, this will be a direct download link or a GitHub release. For now, contact us for early access.]

5 Unzip into a folder you'll remember, like C:\Users\YourName\Documents\SentinelIEP\.

6 Launch the app by double-clicking iep_assistant.py or running from PowerShell:

cd "C:\Users\YourName\Documents\SentinelIEP"
python iep_assistant.py

Quick install (macOS)

1 Install Homebrew if you don't already have it (from brew.sh).

2 Install Python:

brew install python@3.12

3 Install dependencies:

pip3 install faster-whisper sounddevice numpy psutil pillow

4 Download SENTINEL\u00B7IEP, unzip, and run:

python3 iep_assistant.py

Optional: install Ollama for offline AI

SENTINEL\u00B7IEP can run with cloud AI (Anthropic, OpenAI, Google) OR with a local AI model via Ollama. Local mode is the privacy-safe default \u2014 nothing leaves your machine.

1 Download Ollama from ollama.com and install it.

2 Pull a model. The hardware diagnostic that runs on first launch of SENTINEL\u00B7IEP will recommend the best one for your machine. Common picks:

ollama pull llama3.1:8b        # 4.7 GB \u2014 standard
ollama pull qwen2.5:1.5b       # 1 GB \u2014 minimal
ollama pull qwen2.5:14b        # 9 GB \u2014 strong GPU

Optional: configure cloud AI providers

For higher-quality analysis (especially the multi-model case analysis), add API keys for one or more cloud providers in Settings:

Anthropic Claude \u2014 console.anthropic.com
OpenAI \u2014 platform.openai.com
Google Gemini \u2014 aistudio.google.com/apikey

Costs depend on usage. A full multi-model case analysis on a large case file is typically under $1.

Verify the install

After installation, launch SENTINEL\u00B7IEP. You should see:

The first-launch disclaimer dialog \u2014 type your initials and click Continue.
A hardware diagnostic that detects your CPU, RAM, and GPU.
A model recommendation dialog \u2014 click Use to accept the suggested model.
The main application window with the "SENTINEL\u00B7IEP" branding.

Next: your first launch.

Common install problems

"pip is not recognized as an internal or external command"

Python isn't on your system PATH. Try python -m pip install ... or py -m pip install .... If neither works, reinstall Python and check the "Add Python to PATH" box.

"ModuleNotFoundError: No module named 'sounddevice'"

The dependencies didn't install correctly. Run pip install sounddevice numpy faster-whisper psutil pillow again. If you have multiple Python versions, make sure pip and python point to the same one (where python and where pip on Windows; which python and which pip on macOS/Linux).

"The app launches but no audio is captured"

Grant microphone permission to Python (or to Terminal/PowerShell if you launched from there) in your OS settings. On macOS: System Settings \u2192 Privacy & Security \u2192 Microphone.

Still stuck? See Troubleshooting or contact us.