Anthropic to release voice mode, Deepgram launched Aura 2, and more

Voice AI weekly digest

Top Updates 💪

• Anthropic is reportedly working on a voice mode for Claude AI (The Verge)

• Introducing Aura 2: Cost-effective, enterprise TTS (Deepgram)

• Telli, a YC alum, raises pre-seed funding for its AI voice agents (TechCrunch)

• Mango AI launches free tool for realistic voice cloning (MarketWatch)

• Deploying a new interpreter agent at Microsoft (Microsoft Inside Track)

• TicNote: A tiny voice AI assistant (ZDNet)

• HubSpot to snap up Dashworks, bolster its Breeze portfolio (CX Today)

• IBM releases Granite-3-3-8B, a new STT model (MarkTechPost)

• Gong claims most AI agents are “unrealistic or uninspired” (CX Today)

• PolyAI unveils Agent Studio (Smart Customer Service)

• Powered_by Agency’s Virtual SE redefines pre-sales engineering (PR Newswire)

• EU bans AI assistants from virtual meetings (TechRepublic)

• Assort Health raises $26M for GenAI patient call platform (Pulse 2.0)

• Loti AI raises $16.2M in Series A funding (Finsmes)

• Sonix raises CHF 1.8M in funding (Finsmes)

• AI that generates sound from anything (Hackster.io)

• Cisco paves the way with agentic AI collaboration (Zawya)

• How voice AI is revolutionising healthcare efficiency (BW Healthcare World)

• iRocket debuts VoxTalker: Free AI voice generator with 3,200 voices (FOX40)

• iFlytek’s smart translator breaks language barriers (TechBullion)

• Donatos Pizza deploys AI voice ordering to boost sales (OmniTalk Blog)
  

---

Voice AI Podcast 🎙️

In case you missed the latest episode of Voice AI Podcast…

Engineering Corner 😎

• The best translator apps for 2025 (PCMag)

• Free AI chatbot Easemate: a smart AI assistant for you (PCWorld)

• Start building voice intelligence with AssemblyAI’s STT model from AWS Marketplace (AWS Blog)

• Get started with Azure OpenAI’s advanced audio models (Microsoft Foundry)

• A vector quantized masked autoencoder for audiovisual speech emotion recognition (ScienceDirect)

• Advancing Arabic ASR through large-scale weakly supervised learning (arXiv)

• A comprehensive survey of speech summarization (arXiv)

• Common voice API integration mistakes (MarTech Zone)

• 11x.ai: Building AI employees (11x.ai)