Familiar, formerly MetaVoice, builds a duplex speech model for outbound sales calls. It listens while it speaks, so prospects can interrupt mid-sentence, and it runs in your own cloud.
Familiar: Duplex Speech Model for Sales Calls Familiar, formerly MetaVoice, is a San Francisco company building a duplex speech model for outbound sales calls. The model listens while it speaks and reasons over speech rather than text, so a prospect can interrupt it mid-sentence. It is trained on calls from top-performing sales reps and runs in the customer's own cloud. Key Features Duplex Speech Model: Built in-house, it listens and speaks at the same time and replies in 350 ms, with no ASR, TTS, or turn detector to wire up. Trained on Sales Calls: Learned from top-performing reps, it mirrors the prospect's tone and pace, runs discovery, and handles objections. Guardrails and Debugging: Block a response before the caller hears it, and inspect every turn in text and audio. Runs in Your VPC: The model runs in your own cloud, so call data stays with you. Improves Over Time: Coach the agent or fine-tune it on your own calls. Use Cases Lead Qualification: Qualify new sales leads over the phone. Lead Re-Engagement: Win back leads that other voice agents lost. Custom Sales Agents: Build a sales voice agent with your own workflow, tools, and personality.
Speech-to-Speech
HQ COUNTRY
San Francisco, United States
FOUNDED
2022
Is this your company? To claim this listing or update its details, contact us at sunil@uniocommunity.com
