GoKawiil
Tech News
Granola, best known for automated meeting notes, launched a dictation tool that lets Mac users press the Fn key to speak and have text inserted into any app, from emails to prompts. The company says the feature offers unlimited words for free and will expand to iOS and Windows soon. The move mirrors Wispr, which added a meeting notetaker months after building its name on dictation.
Deepgram announced Thursday it has acquired Daily, the company behind the widely used open-source voice AI framework Pipecat, which manages real-time voice and video pipelines including speech-to-text, LLMs, and text-to-speech. Financial terms were not disclosed, but both companies said Pipecat will remain open-source. Daily had raised over $62 million from investors including Tiger Global and Nat Friedman, while Deepgram became a unicorn earlier this year after a $130 million Series C round.
Nvidia released Nemotron 3.5 Transcribe, a free, open-weight transcription model supporting up to 40 languages, alongside version 2 of its open-source Granary speech dataset. The company says the 30-billion-parameter model can translate 31 languages to and from English, process recordings up to an hour long, distinguish up to eight speakers, and run on a single chip ranging from Jetson AGX Thor to Blackwell datacenter hardware. Nvidia claims it outperforms rivals on benchmarks including Hugging Face's Open ASR Leaderboard, though it only works on recorded audio rather than live streams.
CNBC profiled customer service workers whose daily tasks have shifted as companies deploy AI tools for note-taking, scheduling and answering queries. One credit card company worker said AI frees her to focus more on customers, while a telecom worker in Cardiff described AI-generated notes as frequently inaccurate, forcing her to rewrite them herself. Other workers described being pushed to use chatbots for answers and said AI booking agents often routed only the hardest cases to human staff, increasing burnout.
Block, led by Jack Dorsey, has released an AI model called Dessa that processes spoken language directly rather than converting speech to text first. The company says this approach lets the model respond to vocal tone and timing more naturally, similar to how infants learn language before learning to read or write.
Amazon has rebranded its Kindle Unlimited subscription service as Kindle+, keeping the same $11.99/month price and feature set while expanding its audiobook catalog. The expanded library will include more traditionally narrated titles as well as AI-narrated audiobooks produced through Amazon's Virtual Voice tool. Subscribers can still borrow up to 20 ebooks, audiobooks, comics, or magazines at a time across Kindle apps and devices.
At the HumanX conference, PolyAI CTO Shawn Wen said that although full-duplex voice models now exist, reasoning speed must improve further before conversations feel truly natural. Otter CMO Alex Gay added that speaker identification, intent capture and emotionally expressive synthetic voices are needed before AI can serve as digital stand-ins in meetings.
ElevenLabs completed a $300 million tender offer letting employees sell vested shares at a $22 billion valuation, double the $11 billion mark it hit in February when it raised $500 million. The sale was co-led by institutional investors Wellington and T. Rowe Price, and marks the company's second such secondary transaction after a $100 million tender at $6.6 billion in September 2025.
DetectifAI, a San Francisco startup founded by Tarini Padmanabhuni, has built AI models designed to run directly on smartphones to flag AI-generated voices in real time during calls and voice messages. Padmanabhuni started the company after her grandfather was scammed out of money by a deepfake voice impersonating his brother's kidnapping. The company is pitching its software to phone manufacturers to embed detection directly into device operating systems rather than relying on cloud-based analysis.
Boston startup Modulate has closed a $25 million funding round led by Future Ventures, with Hyperplane and Lakestar also participating. The company, founded in 2017 by Mike Pappas and Carter Huffman, runs over 100 models split between signal extraction (detecting emotion, tone, and synthetic voices) and analysis models that assess conversational intent and policy compliance for enterprise voice agents.
In an interview at the Nrth conference in Toronto, ElevenLabs co-founder and CEO Mati Staniszewski discussed the AI voice company's growth, saying it is pacing at $600 million in annual recurring revenue and is reportedly valued at $22 billion by investors, four years after founding. He said the company powers phone support for clients including Klarna, Deutsche Telekom, Cisco, and Adobe, and increasingly competes with businesses that once trained on its technology, such as Decagon.
Iceland-based Treble, founded in 2020 by acoustic engineers Finnur Pind and Jesper Pedersen, has closed an $18 million extension to its Series A round led by Paladin Capital Group, with backing from KOMPAS VC, Frumtak Ventures, EIC, and Omega ehf. The funding brings Treble's total raised to over $40 million, following a $12 million round in 2024, and the company already counts Amazon and Logitech among its customers.