Modulate @modulate_ai

The voice intelligence company | MIT-alum founded. LLMs read text. We listen and understand: Try Velma 2.0. modulate.ai/velma Boston, MA Joined August 2017

Tweets

807
Followers

2K
Following

454
Likes

727

Modulate @modulate_ai

12 hours ago

AI agents are starting to make phone calls. What happens when a business refuses to talk to them? This Tuesday, join Modulate for a LinkedIn Live on detecting AI callers and routing humans to humans, bots to bots. Join us: bit.ly/47njIOu

0 0 0 168 1

View Details

Modulate @modulate_ai

3 days ago

Read the docs and grab 1000 free API credits here: shorturl.at/oHgKp

1 0 0 52 0

View Details

Modulate @modulate_ai

3 days ago

Arabic. English. Arabic again. When a conversation switches languages, Modulate Transcribe detects the language of every utterance, so each turn comes back in the language actually spoken. → 100 languages, Arabic included → Speaker labels that hold across language switches → Batch and streaming

2 0 0 199 1

View Details

Modulate @modulate_ai

4 days ago

Start building with our APIs. 1000 free credits on us: shorturl.at/BVWq8

0 0 1 24 0

View Details

Modulate @modulate_ai

4 days ago

Clinical audiology sessions are messy 👀 Overlapping speakers. Emotional nuance flat transcripts miss. Notes that need to be instant, not "eventually." One developer needed transcription that could keep up. So they built their AI Scribe SaaS on Velma 2. The result: 👉 Live diarization 👉 Emotional context per utterance 👉 Clinical notes generated in real time Their words, not ours: ⭐ 5/5: "Accurate Transcriptions with Emotional Context at a Fantastic Price." This is what audio-native voice intelligence looks like in production 💯

1 0 3 402 0

View Details

Modulate @modulate_ai

5 days ago

Stop transcribing. Start listening: modulate.ai/?utm_source=li…

0 0 0 29 0

View Details

Modulate @modulate_ai

5 days ago

Same words. Five different meanings. “I’m fine” can be relief, exhaustion, sarcasm, or a cry for help. Your transcript can’t tell the difference. Velma can. Audio-native voice intelligence, in real time. Built for teams that need to know what was meant, not just what was said.

1 0 1 107 0

View Details

Carter Huffman @whuffman

6 days ago

Building voice agents is half the problem. Knowing what’s actually happening in those conversations is the other half. In Atlanta for VON, demoing @modulate_ai at VoiceAI LIVE! today alongside @SoundHound and others. 2:40 PM, Room C. Come say hi. I’ll be listening to how you say it.

1 2 2 118 0

View Details

Modulate @modulate_ai

7 days ago

Last week at #SFTechWeek 🎙️ From a rooftop happy hour with VoiceRun to a hackathon with fellow builders, we spent the week connecting with the people shaping the future of voice AI. Thanks to everyone who joined us. Until next time, SF!

2 0 7 809 0

View Details

Carter Huffman @whuffman

a week ago

I used to teach spacecraft to see. Now I teach software to listen. @modulate_ai recently raised $25M, led by Future Ventures, to build the listening layer for AI. Cosmology → JPL → voice. Here’s the path 🧵

1 2 4 125 0

View Details

Modulate @modulate_ai

a week ago

Apply here: modulate.ai/careers?utm_so…

0 0 1 80 0

View Details

Modulate @modulate_ai

a week ago

We're hiring. 3 roles, 1 mission: make voice a force for good online. → Backend Engineer → Staff Backend Engineer → Senior ML Engineer Somerville, MA. Full time. Hybrid.

6 2 8 2K 1

View Details

Modulate @modulate_ai

2 weeks ago

1000 free API credits here: shorturl.at/BVWq8

0 0 1 46 0

View Details

Modulate @modulate_ai

2 weeks ago

Most developers don't switch speech-to-text providers for a small win. They switch when one option is obviously, undeniably better 🏅 That's what happened here: ⭐ 5/5 "Outstanding Accuracy, Speed, and Value" Simple API. Clear docs. Integration that takes minutes, not days. Start building.👇

1 0 1 99 0

View Details

Modulate @modulate_ai

2 weeks ago

Speech recognition = audio in, text out. That's the easy version. Our CEO and Co-founder @whuffman on why real understanding means reading accent, tone, emotion, and cadence too - not just the words. We're hiring for people who think this way. See open roles: hubs.ly/Q04xDKp90 #hiring

0 1 9 400 0

View Details

Steve Jurvetson @FutureJurvetson

3 weeks ago

Winnowing nuance from noise Modulate runs over 100 voice-native models to discern emotion, tone, intent, and language, and detect synthetic and deepfake voices (ranked #1 on 🤗). We just led their seed round; news today: techcrunch.com/2026/09/28/mod… Product: modulate.ai/velma

Modulate @modulate_ai

3 weeks ago

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

3 3 22 224K 7

5 5 30 6K 2

View Details

Hasan Toor @hasantoxr

7 months ago

🚨 A startup just hit #1 on @huggingface for deepfake audio detection. 98.9% accuracy. 30–1000× cheaper than every competitor. Nobody is talking about this. It's called Velma by @modulate_ai and it changes everything about voice security.

20 29 69 15K 57

View Details

Michael Pappas @mpappas74

4 months ago

x.com/i/article/2069…

3 1 11 148K 0

View Details

FELIX @FellMentKE

7 months ago

🚨 BREAKING: Transcription just got 90x cheaper. One of the biggest hidden costs for AI startups is Speech-to-Text. Velma-2 by @modulate_ai just launched: $0.03 per audio hour transcribe. Here’s why this changes everything for AI founders:

24 37 60 159K 64

View Details

Modulate @modulate_ai

2 weeks ago

Velma is designed for real-world audio, with evaluation across 14 speech deepfake datasets, including varied accents and challenging audio conditions. On deepfake detection, Velma achieves 98.9% accuracy / 1.1% EER and ranks #1 on the @huggingface Speech Deepfake Arena. We’re seeing just 11 false positives per 1K real calls in our benchmark - an important metric for production use.