# agituts/gemini-2-podcast architecture

> Architecture diagram of the GitHub repository agituts/gemini-2-podcast, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2024-12-27). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/agituts/gemini-2-podcast
- Repository: https://github.com/agituts/gemini-2-podcast

## Overview

##### Component Details and Relationships

a) Input Handler
- Rectangle labeled "Content Ingestion"
- Connection to Script Generator

b) Script Generation (generate_script.py)
- Connection to Gemini 2.0 API
- System instructions input (system_instructions_script.txt)

c) Audio Generation (generate_audio.py)
- Connection to Google TTS API
- System instructions input (system_instructions_audio.txt)
- Two parallel processes for Voice A and Voice B

d) Audio Processing (audio_processor.py)
- Rectangle for audio processing module
- FFmpeg integration
- Audio mixing and formatting operations

##### Data Flow Representation

**Input Sources → Content Ingestion → Script Generation → Audio Generation → Audio Processing → Final Output**

##### External Dependencies

External services in clouds:
- Google Gemini 2.0 API
- Google Text-to-Speech API
- FFmpeg
- PortAudio

##### Configuration Elements

Configuration components:
- Environment variables (.env)
- System instruction files
- Requirements and dependencies

##### Technical Details

Annotations for:
- Audio specifications (24000Hz, 16-bit, stereo)
- API integration points
- File format transformations
- Error handling and retry mechanisms

## Diagram (Mermaid)

```mermaid
flowchart TB
    %% Styles
    classDef input fill:#a8d1f0,stroke:#333
    classDef processing fill:#90EE90,stroke:#333
    classDef external fill:#FFE4B5,stroke:#333
    classDef audio fill:#FFA07A,stroke:#333
    classDef config fill:#DDA0DD,stroke:#333
    classDef output fill:#98FB98,stroke:#333

    %% Input Layer
    subgraph InputSources
        PDF["PDF Files"]:::input
        URL["URLs"]:::input
        TXT["Text Files"]:::input
        MD["Markdown"]:::input
    end

    %% Configuration
    subgraph Config
        ENV[".env File"]:::config
        REQ["Requirements"]:::config
        click REQ "https://github.com/agituts/gemini-2-podcast/blob/main/requirements.txt"
    end

    %% Main Pipeline
    ORCH["Pipeline Orchestrator"]:::processing
    click ORCH "https://github.com/agituts/gemini-2-podcast/blob/main/generate_podcast.py"

    %% Content Processing
    CI["Content Ingestion"]:::processing

    %% Script Generation
    subgraph ScriptGen
        SG["Script Generator"]:::processing
        click SG "https://github.com/agituts/gemini-2-podcast/blob/main/generate_script.py"
        SI["System Instructions"]:::config
        click SI "https://github.com/agituts/gemini-2-podcast/blob/main/system_instructions_script.txt"
    end

    %% External Services
    GAPI["Gemini 2.0 API"]:::external
    TAPI["Google TTS API"]:::external
    FFMPEG["FFmpeg"]:::external

    %% Audio Generation
    subgraph AudioGen
        AG["Audio Generator"]:::audio
        click AG "https://github.com/agituts/gemini-2-podcast/blob/main/generate_audio.py"
        AI["Audio Instructions"]:::config
        click AI "https://github.com/agituts/gemini-2-podcast/blob/main/system_instructions_audio.txt"
        VA["Voice A Process"]:::audio
        VB["Voice B Process"]:::audio
    end

    %% Audio Processing
    AP["Audio Processor"]:::audio
    click AP "https://github.com/agituts/gemini-2-podcast/blob/main/audio_processor.py"

    %% Final Output
    OUT["Final Podcast Audio"]:::output

    %% Connections
    PDF & URL & TXT & MD --> CI
    CI --> ORCH
    ORCH --> SG
    SG <--> GAPI
    SI --> SG
    SG --> AG
    AG <--> TAPI
    AI --> AG
    AG --> VA & VB
    VA & VB --> AP
    AP <--> FFMPEG
    AP --> OUT
    ENV --> ORCH
    REQ --> ORCH

    %% Legend
    subgraph Legend
        L1["Input Sources"]:::input
        L2["Processing Components"]:::processing
        L3["External Services"]:::external
        L4["Audio Components"]:::audio
        L5["Configuration"]:::config
        L6["Output"]:::output
    end
```
