Liberate Lands $50M for Voice AI Agents

October 21, 2025

🔊 Soundcheck

  • Liberate Lands $50M for Voice AI Agents

  • WellSaid Boosts Enterprise AI Voice Quality

  • Cookiy AI’s Agentic Voice AI Gets $7M+ Boost

  • Local TTS with Dia‑1.6B

Read time: 4 minutes

🔥 Hot Mic

Big moves, deep dives, and standout stories.

Liberate secures $50 million Series B led by Battery Ventures to expand its AI agents across insurance workflows.

Liberate, a San Francisco–based insurtech startup specializing in voice AI, has raised $50 million in a Series B round led by Battery Ventures. This brings its total funding to $72 million and values the company at around $300 million post-money.

The firm builds AI reasoning agents that automate full insurance workflows—sales, servicing, claims—across voice, email, SMS, and digital channels. Its agents integrate with carriers’ core systems, enabling end‑to‑end task resolution with auditability and human‑in‑the‑loop safeguards.

Liberate’s platform has shown impressive results: 263 % ROI for a major insurer, 15 % sales boost, 23 % cost reduction, and claim response time cut from 30 hours to 30 seconds. The new funding will help deepen its AI reasoning capabilities and drive wider deployment across insurers and agencies.

Battery Ventures’ Marcus Ryu, who will join Liberate’s board, and other investors like Canapi Ventures, Redpoint, Eclipse, and Commerce Ventures backed this round, reflecting strong confidence in liberate’s mission.

Key Points:

  • Series B raise: $50 million led by Battery Ventures

  • Total funding now $72 million

  • Valuation estimated at $300 million post-money

  • AI agents automate entire insurance workflows across channels

  • Documented 263 % ROI, 15 % sales increase, 23 % lower costs

  • Hurricane claim response dropped from 30 hours to 30 seconds

  • Reinforcement‑learning AI with audit and human‑in‑loop safeguards

  • New board member: Marcus Ryu from Battery Ventures

Takeaway: Liberate’s funding round signals growing investor trust in deeply integrated, reasoning‑based AI agents that tackle end‑to‑end insurance tasks—delivering fast ROI while scaling across tightly regulated workflows.

WellSaid Labs launches “Caruso” enhancements—faster, highly natural AI voices with compliance tools built for enterprise.

WellSaid Labs just revealed its latest AI voice platform overhaul, centered around the powerful new “Caruso” model and intelligent “AI Director” features. Enterprises now get lifelike speech at unprecedented speed and fidelity. The updates also bring word‑level tuning, instant previews, and nearly broadcast‑ready audio at up to 96 kHz.

On the compliance side, WellSaid has doubled down on enterprise needs—adding roles for collaborators, billing admins, SSO, language coverage, and audit support. The platform remains firmly grounded in ethical AI, using licensed actor voices only, with SOC 2 and GDPR compliance baked in, making it a trusted choice for highly regulated industries.

Key Points:

  • Caruso model delivers AI voice rendering about 30 % faster.

  • AI Director enables precise emotional, pitch, and pace control.

  • Studio outputs high‑fidelity audio up to 96 kHz.

  • Enterprise features include SSO, admin roles, and audit tools.

Takeaway: WellSaid Labs isn’t just improving voice synthesis speed and realism—it’s building a secure, controllable, and compliant platform that enterprises can trust to scale synthetic audio content reliably.

Cookiy AI raised an oversubscribed $7M+ pre-seed round to scale its Agentic Voice AI for real-time consumer insights.

Cookiy AI, founded by former Instagram and TikTok executive Davin YC Dong, just closed an oversubscribed pre-seed round exceeding $7 million. The funding, led by Liquid2, Converge, GoAhead, and UpHonest, underscores growing excitement around Agentic Voice AI.

The platform transforms large-scale voice conversations into structured insights in under 48 hours, promising faster, cheaper, and more empathetic consumer research. It aims to democratize deep consumer understanding for brands and agencies, shifting research from reactive to proactive.

Key Points:

  • Raised over $7M in oversubscribed pre-seed round.

  • Backed by Liquid2, Converge, GoAhead and UpHonest.

  • Agentic Voice AI turns conversations into structured insights.

  • Delivers decision-ready consumer data in under 48 hours.

Takeaway: Cookiy AI’s successful raise spotlights the appeal of agentic voice-powered consumer research, offering brands a faster, more humane way to tap into customer motivation and preferences.

A hands-on guide to installing and running Dia‑1.6B locally for private, expressive TTS.

This guide walks you step by step through setting up Dia‑1.6B, an open‑source TTS model by Nari Labs, on your own machine. It covers prerequisites, installation, running the model via a Gradio interface, and generating expressive dialogue with non‑verbal cues. You’ll also find tips for voice cloning, customization, and future updates like CPU support and quantized versions to reduce hardware requirements.

Key Points:

  • Dia‑1.6B is an open‑source dialogue‑focused TTS model with 1.6B parameters.

  • Generates multi‑speaker dialogue using tags like [S1] and [S2].

  • Adds natural non‑verbal sounds such as (laughs) or (coughs) from text cues.

  • Requires a CUDA‑enabled NVIDIA GPU with ~10 GB VRAM; CPU support planned.

Takeaway: Dia‑1.6B brings expressive, private, locally runnable TTS within reach—offering full control, voice cloning, and realistic dialogue generation without cloud dependencies.

🎙️ Mic Drop

What else is making noise in voice AI.

WellSaid Labs touts new enterprise-grade TTS offering secure, high-quality, human-like audio for regulated and compliance-focused industries. (businesswire.com)

WellSaid Labs pivots to secure, enterprise-first voice AI, launching Caruso model and redesigned Studio targeting regulated sectors. (geekwire.com)

Fundamento will use proceeds to enter international markets and expand BFSI-focused voice AI solutions for collections and profiling. (economictimes.indiatimes.com)

Barcelona-based SLNG secures funding to deliver multilingual, global speech AI infrastructure aimed at diversifying market options. (eu-startups.com)

SLNG.ai aims for a developer-first, compliance-ready voice AI platform supporting non-English markets and global deployment. (tech.eu)

SoundHound secures leadership status for Conversational AI as Amelia 7 powers real-world contact center transformation. (simplywall.st)

Oracle launches new voice control integrations for real-time data access and communication in public safety scenarios. (stocktitan.net)

CX Today explores the impact of voice AI on operational costs and efficiency in modern contact centers. (cxtoday.com)

Tata Communications launches agentic, speech-to-speech voice AI platform for banking and financial service customer engagement. (futurecio.tech)

Dstny integrates legacy PBX with AI voice, improving workplace communication for enterprise clients navigating unified comms shifts. (uctoday.com)

Microsoft adds 'Hey Copilot' voice assistant to Windows 11, reflecting OS-level commitment to voice-first user interaction. (ummid.com)

WideBot becomes first MENA firm offering WhatsApp voice bots, expanding conversational AI to new service environments. (disruptafrica.com)

AiBuddy debuts a unified CRM leveraging voice AI for streamlined customer relationship management and workflow automation. (markets.financialcontent.com)

Feature spotlights emerging Indian voice AI startups building culturally relevant, multilingual speech technologies. (analyticsindiamag.com)