ELI5 Anything 🧠

A local AI tool that explains any concept at four reading levels — from a 5-year-old to a domain expert — using Ollama. Runs entirely on your machine. No API keys. No internet required.

Demo

📝 Enter concept: RBAC
============================================================
  Concept: RBAC
============================================================
🧒 ELI5 (Explain Like I'm 5)
----------------------------------------
Imagine a school where only teachers can go into the supply closet,
only the principal can enter the server room, and students can only
go to classrooms. RBAC is like that — everyone gets access to only
the places they're supposed to be in, based on their job.

🎒 High School
----------------------------------------
RBAC stands for Role-Based Access Control. Instead of giving each
person individual permissions, you assign them a "role" (like Admin,
Editor, or Viewer), and that role comes with a preset list of things
they can do. It's like how a part-time employee badge only opens
certain doors, while a manager badge opens more.

🎓 College Level
----------------------------------------
...

🔬 Expert
----------------------------------------
...

Features

• 4 reading levels: ELI5, High School, College, Expert
• Interactive mode: run it and keep asking questions
• Single-shot mode: pass a concept directly as an argument
• Auto-detects your installed Ollama model
• Fully local: no data leaves your machine
• Zero dependencies: pure Python standard library

Requirements

• Python 3.10+
• Ollama installed and running
• A pulled model (e.g. ollama pull llama3.2)

Installation

1. Install Ollama

# macOS (Homebrew)
brew install ollama

# or download the app at https://ollama.com/download

2. Start Ollama and pull a model

ollama serve          # start the local server
ollama pull llama3.2  # download a model (~2GB)

3. Clone this repo

git clone https://github.com/debugbae/eli5-anything.git
cd eli5-anything

No pip install needed — zero third-party dependencies.

Usage

# Interactive mode — keeps prompting until you quit
python eli5.py

# Explain a specific concept (all 4 levels)
python eli5.py "transformer architecture"

# Single level only
python eli5.py --level eli5 "machine learning"

# Multiple specific levels
python eli5.py --level hs --level college "zero trust security"

# Use a specific Ollama model
python eli5.py --model mistral "quantum computing"

Why I Built This

Jargon is a barrier. Whether you're onboarding a new team member, explaining a technical concept to an executive, or just trying to understand something yourself — having a tool that can instantly translate complexity into clarity is genuinely useful.

This project also demonstrates:

• Structured prompt engineering across multiple personas
• Local LLM integration via the Ollama API
• Clean CLI design with argparse
• Graceful error handling for API connectivity issues

How It Works

Each reading level has a carefully crafted system prompt that instructs the model to adjust vocabulary, depth, and framing. The tool sends the user's concept + the level prompt to Ollama's local REST API and streams back the response.

User input → Level prompt + Concept → Ollama API → Formatted output

Levels

Level	Audience	Style
eli5	5-year-old	Simple analogy, 2–3 sentences
hs	High schooler	Relatable example, some terms defined
college	Undergrad	Precise, includes mechanism + significance
expert	Practitioner	Technical, nuanced, edge cases