We just raised $25M in new funding led by Future Ventures (@VenturesVc), with @HyperplaneVC and @lakestarvc - bringing Modulate’s total funding to $60 million. Building the next phase of audio-native AI: more models, more developers, more of the stack that actually understands voice. Where we stand today: - Ranked #1 on Hugging Face for deepfake detection and transcription - 600M+ hours of audio analyzed - 10M+ hours analyzed every month A foundational layer of the AI stack is getting more powerful. More soon.

Founders get told to ignore the doubters. So do early employees. Our co-founder Mike: on knowing when to trust your gut vs. when the "obvious" advice is actually right. We're looking for people who can tell the difference 🧠 : hubs.ly/Q04xDKsr0 #hiring

Last week, we joined the developer community at @WeAreDevs in San Jose 🙌 Great conversations, new connections, and plenty of discussions around what’s next for AI, voice, and the developer ecosystem.

"A lot of people assumed we were on our way to being acquired. That's never been the mission." @modulate_ai, Co-founder Mike Pappas on why the company was built to go the distance - not to get bought. 👇

شركة @modulate_ai الأمريكية تجمع 25 مليون دولار لتطوير تقنيات ذكاء اصطناعي تكشف الأصوات الاصطناعية والتزييف العميق في المكالمات. تخطط الشركة لاستخدام التمويل في تعزيز قدرتها على مواجهة الاحتيال الصوتي ودعم قواعد الامتثال في القطاعات المنظمة. jawlah.co/65301

BREAKING 🚨 @modulate_ai, cofounded by @whuffman @mpappas74, raised $25M in new funding led by Future Ventures @VenturesVc with @HyperplaneVC and @lakestarvc, to expand its audio-native AI models and developer ecosystem Learn more: fondo.ai/47qiNfW

Building voice intelligence that understands what’s happening beyond the transcript. Excited to have Future Ventures, Hyperplane, and Lakestar backing us as we build what’s next 🚀

Most fraud tools are reading the wrong thing. 👀 They monitor transactions. They flag metadata. They analyze the transcript after the call ends. All useful, BUT none of it catches fraud while it's happening, because the tells aren't in the words. They're in the voice. Here's what a transcript misses on every fraud call: 1️⃣ Scripted or rehearsed speech patterns 2️⃣ Emotional incongruence - calm under pressure, urgency without stress 3️⃣ Hesitation, contradictions, and recall gaps 4️⃣ Voice mismatch against the account holder 5️⃣ Synthetic or AI-generated voice indicators Modulate's Velma is built to catch all five live: flagging deceptive intent to your agents before the call completes, not after the damage is done. Account takeover fraud cost US adults $15.6B in 2024, up 23% year over year. 📈 And Velma's deepfake detector generates fewer than half the errors of the next-best model on @huggingface independent leaderboard. Stop reading the transcript. Start listening to the call: hubs.ly/Q04xdbt90

This team built the first AI models that understand voice - not just what's said, but how it's said. In a few hours, we're sharing what's next. 👀

What does it take to make voice AI actually understand human speech? Join Modulate CEO @whuffman and @jeffpulver for a 25-minute conversation on: → How Modulate evolved from gaming + trust & safety into Voice AI → Why understanding voice goes beyond transcription → AI agents, fraud & deepfake detection → The future of voice infrastructure → What VCON could mean for the communications industry Tune in here: hubs.ly/Q04ylZZ90

Five people. A redeye flight. And a moment that made it real. ✨ @mpappas74, Co-founder of Modulate, shares one of the company's most memorable early moments: Early on, Mike and @whuffman flew from Boston to San Francisco and back in a single day to meet prospective customers - landing home on a redeye at 9am, exhausted, but founders don't get to clock out. They trudged into the office anyway. And when they opened the door, their first employees were already at the whiteboard, deep in an animated discussion, solving problems on their own. That was the moment it stopped being just the two of them - and started being a team, a company, something bigger than its founders. Today, this team is expanding. We're hiring for multiple roles: hubs.ly/Q04xDKyy0 #hiring

We’re at @WeAreDevs in San Jose and having a GREAT time 🚀 The energy here has been unreal. If you’re at #WeAreDevelopers, come find us!

You can't just dump recordings into a model and expect it to work. Modulate CEO @whuffman on why voice AI lives and dies on data quality, not data quantity - and why context (language, region, setting) changes everything. 👇 hubs.ly/Q04xDKzs0

"Hi, this is Sarah calling to update my account details." Correct name, verified details, calm delivery - nothing in the words raises a flag. But the voice carries signals text never will: 👉 Synthetic voice indicators 👉 Emotional incongruence 👉 Rehearsed speech cadence 👉 Elevated stress markers Velma flagged all 4 in real time, each one tied back to the original audio for review. The synthetic-voice signal is the one that matters most as cloning gets cheaper. Velma Deepfake Detect, according to @huggingface Speech Deepfake Arena leaderboard, ranks #1 with a 1.1% equal error rate and 98.9% accuracy — identifying a cloned voice from as little as 2.5 seconds of audio 💯 It also doesn't stop checking after the first few seconds. Velma scores every segment of a call continuously, so a fraudster who opens with a real voice and switches to a clone mid-call gets caught too. See the model: hubs.ly/Q04xd8cH0