memoair-voice 0.3.2

Local-runtime voice memory client for MemoAir (sits on top of memoair-runtime)

memoair-voice

Local-runtime voice memory client for MemoAir. Wraps the loopback voice-runtime sidecar, keeps per-user memory on-device, syncs to cloud.

Install

pip install memoair-voice

Quick start

from memoair_voice import MemoAirVoiceAgent

agent = MemoAirVoiceAgent(
    api_key="memoair_pk_...",        # account-scoped API key
    project_id="proj_xxx",            # which customer project
    agent_id="agent_xxx",             # which agent identity
    user_id="caller_phone_hash",      # optional, per-call end user
)

# Or pull from MEMOAIR_API_KEY / MEMOAIR_PROJECT_ID / MEMOAIR_AGENT_ID /
# MEMOAIR_USER_ID (+ optional MEMOAIR_BASE_URL):
agent = MemoAirVoiceAgent.from_env()

# Headers carrying every identity dimension:
#   Authorization: Bearer memoair_pk_...
#   X-Project-Id: proj_xxx
#   X-Agent-Id: agent_xxx
#   X-User-Id: caller_phone_hash
agent.headers()

# Build a runtime-backed client:
async with agent.client() as client:
    await client.search_memory("what time is standup")
    await client.save_response(
        user_text="What time is standup?",
        assistant_text="10:30 IST.",
    )

Docs

Full documentation at https://docs.memoair.dev/voice