Skip to content

README.md

Latest commit

 

History

History
163 lines (112 loc) · 3.67 KB

README.md

File metadata and controls

163 lines (112 loc) · 3.67 KB
Raw

🧠 MasterSpeakAI

MasterSpeakAI is a GenAI-powered speech analysis app with:

  • 🎛️ A Streamlit frontend using st.navigation and st.Page
  • 🔥 FastAPI backend
  • 🧪 GPT-powered analysis (via OpenAI)
  • 💾 Local SQLite databases for storing analyses, prompt presets, and model settings
  • ⚙️ Model-specific configuration (e.g., temperature, max_tokens)
  • 📦 Managed with Poetry

🚀 Features

  • 🎙️ Analyze speech content using OpenAI GPT models (3.5, 4, 4o, etc.)
  • 🧠 Create and edit reusable system prompt presets
  • ⚙️ Configure model settings (e.g., temperature) per model
  • 💾 View and retrieve records using compact scrollable DB viewers
  • 📎 Manual ID lookup for speech/prompt/model records
  • 🧩 Grouped sidebar navigation for logical app structure

📁 Project Structure


kaiser-data-masterspeakai\_streamlit/
├── backend/
│   ├── database.py              # SQLModel tables for speech, prompts, models
│   ├── main.py                  # FastAPI app with /analyze endpoint
│   ├── init\_db.py               # Creates all DB tables
├── frontend/
│   ├── app.py                   # Main entrypoint with st.navigation sidebar
│   └── pages/
│       ├── 0\_Home.py
│       ├── 1\_Speech\_Analysis.py
│       ├── 2\_Database\_Viewer.py         # Speech DB (manual ID lookup)
│       ├── 3\_Prompt\_Settings.py         # Create/edit prompts
│       ├── 4\_Prompt\_Viewer.py           # Prompt DB (manual ID lookup)
│       ├── 5\_Model\_Settings.py          # Edit model config
│       └── 6\_Model\_Settings\_Viewer.py   # View saved configs
├── init\_db.py                  # Calls backend.init\_db functions
├── pyproject.toml              # Poetry config
├── requirements.txt            # pip fallback
├── speech\_db.db
├── prompt\_db.db
├── model\_settings.db
└── .env                        # For OPENAI\_API\_KEY

🧭 Sidebar Navigation

Organized using st.navigation() with icons and sections:

  • 🧠 MasterSpeakAI

    • 🏠 Home

  • 🔊 Speech Analysis

    • 🎙️ Analyze Speech
    • 💾 Speech DB Viewer (with manual ID input)

  • 🛠️ Prompt Settings

    • 🛠️ Prompt Editor
    • 📜 Prompt DB Viewer (with manual ID input)

  • ⚙️ Model Settings

    • 🧩 Edit Model Settings (temperature, top_p, etc.)
    • 🗃️ Model DB Viewer

📦 Dependency Management with Poetry

This project uses Poetry to manage dependencies and virtual environments.

📥 Install Poetry

curl -sSL https://install.python-poetry.org | python3 -

Ensure it's on your PATH:

export PATH="$HOME/.local/bin:$PATH"

📦 Install Dependencies

git clone https://github.com/your-org/kaiser-data-masterspeakai_streamlit.git
cd kaiser-data-masterspeakai_streamlit
poetry install

🐍 Run the App

poetry run streamlit run frontend/app.py

🔧 Initialize the Databases

poetry run python init_db.py

Creates:

  • speech_db.db
  • prompt_db.db
  • model_settings.db

🧪 Run Backend API

poetry run uvicorn backend.main:app --reload

Visit the docs at: http://localhost:8000/docs

🔐 Environment Variables

Create a .env file with your OpenAI key:

OPENAI_API_KEY=your-key-here

🗂️ Future Improvements

  • ⏳ User authentication
  • ⏳ Role-based prompt/model visibility
  • ⏳ Speech-to-text input support
  • ⏳ Export results to CSV
  • ⏳ Admin edit/delete for DB records

🪪 License

MIT License.