Yentofi

Adaptive Intelligence Companion

Make ordinary people extraordinary.

Getting Started • Web UI • How it Works • Features

---

💡 What is Yentofi?

Yentofi is an open-source adaptive intelligence companion that grows with you.

Unlike generic AI chatbots, Yentofi builds a persistent model of who you are — your knowledge, your gaps, your thinking patterns, your goals — and uses it to push you just beyond your current limits every single session.

The difference: It's not a tutor. It's not a search engine. It's a thinking partner that never lets you stay comfortable.

The gap Yentofi fills: Every major AI product is built for the average user. Yentofi is built for the person who wants to become extraordinary.

---

⚙️ How it Works

graph TD
    A[Speak or Type] --> B[Yentofi Reads Your Profile]
    B --> C[Set Session Goal Above Your Level]
    C --> D[Socratic Questioning]
    D --> E[Extract Knowledge & Update Profile]
    E --> F[Next Session is Smarter About You]

Loading

Your profile lives locally on your machine. Nothing is sent to any server except the conversation text to the LLM API for inference.

---

🚀 Getting Started

1. Clone the repo

git clone https://github.com/rishibagale/yentofi.git
cd yentofi

2. Install dependencies

pip install -r requirements.txt

3. Set up your API key

cp .env.example .env
# Edit .env and add your Groq API key
# Free key at: https://console.groq.com

Mode	Command
Web UI (Mobile)	streamlit run app.py
Text Mode	python main.py
Voice Mode	python main.py --voice

---

📱 Web UI & Mobile

Yentofi now features a beautiful, mobile-responsive web interface.

One-Line Install & Run

If you have Python installed, just run this in your terminal:

pip install -r requirements.txt && streamlit run app.py

Access on Mobile

1. Run the command above on your computer.
2. Ensure your phone and computer are on the same Wi-Fi.
3. Open the Network URL shown in your terminal (e.g., http://192.168.x.x:8501) on your phone's browser.
4. Enter your Groq API Key in the sidebar.

---

✨ Features

• 🧠 Adaptive profiling — Builds a persistent model of your knowledge and thinking style across sessions.
• 🗣️ Socratic method — Asks before it answers, never lectures.
• 📈 Zone of proximal development — Always pushes you just beyond your current level.
• 🎙️ Voice input — Powered by faster-whisper, runs fully on CPU.
• 🔒 Privacy-first — Your profile never leaves your machine.
• ⚡ Runs on 8GB RAM — No GPU required, uses Groq's free API for inference.

---

📂 Project Structure

yentofi/
├── core/
│   ├── agent.py       # Main agent loop and LLM orchestration
│   ├── session.py     # Session management
│   └── prompts.py     # All system prompts
├── memory/
│   ├── profile.py     # User profile model
│   └── store.py       # Persistence layer
├── voice/
│   └── listener.py    # faster-whisper voice input
├── knowledge/
│   └── retriever.py   # RAG pipeline (coming soon)
├── docs/
│   └── architecture.md
└── main.py

---

🗺️ Roadmap

• RAG Pipeline: Curated neuroscience/psychology knowledge base.
• Web UI: Session history and progress visualization.
• Multimodal: Image analysis and visual reasoning.
• Local LLM: Support via Ollama for full offline mode.
• Export: Session insights as markdown files.

---

🤝 Contributing

Contributions are what make the open-source community an amazing place to learn, inspire, and create. Any contributions you make are greatly appreciated.

Read CONTRIBUTING.md to get started.

---

🛠️ Built by

Rishi Bagale github.com/rishibagale

---

“Make ordinary people extraordinary.”

y e n t o f i