GH Repository · xzf-thu
VoiceMem
Infrastructure for the next generation of voice agents, designed to provide universal memory. It is divided into a left brain and a right brain, storing information and emotions respectively, while a fully streaming architecture eliminates latency at the fundamental level.
- stars
- 1,800
- 30-day movement
- +510170/day
- Related entries
- 60
- Connections
- 1
agent-appPythonaiapplicationaudio-streaminggenaiai-agentslong-term-memorymemorymemory-managementllmai-toolspythonvoice-assistantvoice-agent
VoiceMem is a Python-based voice agent application that provides universal memory infrastructure, splitting memory into a 'left brain' for information and a 'right brain' for emotions. It uses a fully streaming architecture intended to eliminate latency at a fundamental level.
Use it when you need a voice agent that remembers both factual information and emotional context across interactions.
Use it to
- Add long-term memory to a voice assistant
- Store emotional context alongside factual memory
- Build low-latency streaming voice agents
- Experiment with dual-brain memory architectures
For Developers building memory-equipped voice agents in Python
- Role
- agent-app
- Language
- Python
- Licence
- Apache-2.0
- Forks
- 126
- Open issues
- 7
- Last push
- 2026-09-14
- Latest release
- v0.0.1 · 2026-09-01
topicsvoice-agentslong-term-memorymemory-managementaudio-streamingllmpython