K2-k2/QuickTalk-AI

🎙️ QuickTalk AI

QuickTalk AI is a high-performance audio intelligence tool that transforms long-form content—like podcasts and videos—into concise, 30-second audio briefs. Built with Streamlit, OpenAI Whisper, and Google Gemini 2.5 Flash, it automates the "listen, summarize, and speak" workflow.

🌟 Key Features

• —Multimodal Input: Support for both video (MP4) and audio (MP3) uploads.
• —Whisper STT: Uses OpenAI's Whisper model to accurately transcribe spoken content.
• —Gemini 2.5 Summarization: Leverages Google's latest Flash model to distill transcripts into three punchy, high-impact bullet points.
• —Voice Briefing: Automatically converts the AI summary back into a high-quality audio file using gTTS.
• —Transcript Access: Toggleable view for the full extracted text.

🚀 Deployment on Hugging Face Spaces

To deploy this successfully, follow these steps:

1. 1.SDK: Select Streamlit.
2. 2.Secret Keys: Add your Gemini API Key in Settings > Variables and Secrets as GOOGLE_API_KEY_N.
3. 3.Requirements: Your requirements.txt must include:

    streamlit
    openai-whisper
    torch
    google-generativeai
    moviepy
    gTTS
    yt-dlp

1. 1.System Packages: Create a packages.txt file in your repository and add:

    ffmpeg

🏗️ The Tech Stack

• —Frontend: Streamlit (UI/UX)
• —Audio Extraction: MoviePy (Video-to-Audio conversion)
• —Speech-to-Text: OpenAI Whisper (Base Model)
• —Intelligence: Google Gemini 2.5 Flash
• —Speech Synthesis: gTTS (Google Text-to-Speech)

💻 Local Setup

# 1. Clone the repository
git clone [https://huggingface.co/spaces/K2-k2/QuickTalk-AI
](https://huggingface.co/spaces/K2-k2/QuickTalk-AI
)

# 2. Install dependencies
pip install -r requirements.txt

# 3. Set your Environment Variable
# (On Windows use 'set', on Mac/Linux use 'export')
export GOOGLE_API_KEY_N="your_api_key_here"

# 4. Run the app
streamlit run app.py

Developed by Kesar Agarwal an aspiring AI Engineer focused on creating accessible, agentic AI tools.

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference