How shreyas850/neel-madhav-ai works
Neel Madhav AI is a local, voice-activated desktop assistant combining conversational models with command execution and spiritual responses. In the main workflow, the desktop launcher loads speech recognition and the logic brain; recognized speech is routed to system, media, scripture, mantra, or chat handling, and the launcher speaks returned responses. A Flask chat endpoint is an alternate entry point. The README also describes dual model personalities and local GPU inference. Additional automation, Google access, and memory utilities exist, but sampled sources do not show them wired into the main workflow.
Main entry points: logic_brain.py Command Router · launcher.py Desktop Launcher · voice_core.py Voice Recognition · brain_loader.py Local Model Runtime
Components 22
Interaction
- Desktop Launcher
launcher.py - Voice Recognition
voice_core.py - Chat API
server.py - Speech Output
logic_brain.py
Intelligence
- Command Router
logic_brain.py - Local Model Runtime
brain_loader.py - Conversational AI
logic_brain.py - Local GGUF Model
Assistant Capabilities
- System Commands
logic_brain.py - Desktop Skills
skills.py - Media Playback
logic_brain.py - Gita Verses
logic_brain.py - Mantra Recitation
logic_brain.py
Additional Services
- ElevenLabs TTS
- YouTube
- Vedic Scriptures API
- Hardware Monitoring
autonomy.py - Google Calendar and Gmail
google_agent.py - Conversation Memory
memory_core.py - JSON Memory File
memory_core.py - Settings Store
db.py
Other components
- User
Connections 18
- User → Desktop Launcher: speaks to
README.md - Desktop Launcher → Voice Recognition: starts listening
launcher.py - Voice Recognition → Command Router: transcribes and calls
launcher.py - Desktop Launcher → Command Router: routes API input
launcher.py - Command Router → Desktop Skills: calls app opener
logic_brain.py - Command Router → Media Playback: routes playback
logic_brain.py - Media Playback → YouTube: plays via
logic_brain.py - Command Router → Gita Verses: routes verse requests
logic_brain.py - Gita Verses → Vedic Scriptures API: requests verses
logic_brain.py - Command Router → Mantra Recitation: routes chant requests
logic_brain.py - Command Router → Conversational AI: routes other input
logic_brain.py - Conversational AI → Local Model Runtime: queries model
logic_brain.py - Local Model Runtime → Local GGUF Model: loads and runs
brain_loader.py - Command Router → Desktop Launcher: returns response
launcher.py - Desktop Launcher → Speech Output: speaks response
launcher.py - Speech Output → ElevenLabs TTS: uses when configured
logic_brain.py - Hardware Monitoring → User: speaks alerts
autonomy.py - Conversation Memory → JSON Memory File: reads and writes
memory_core.py