LiveKit raises $45M, Amazon unveils Nova Sonic, and more

Voice AI weekly digest

Top Updates 💪

• LiveKit raises $45M Series B to build a platform for voice AI agents (LiveKit Blog)

• Amazon unveils a new AI voice model, Nova Sonic (TechCrunch)

• Google Docs adds NotebookLM-style features (Yahoo Tech)

• Aircall unveils AI voice agent for growing businesses (Business Wire)

• Google’s Customer Engagement Suite enhances customer experience (Info-Tech)

• Observe.AI redefines the future of contact centres (Inc42)

• Alorica unveils advanced conversational AI evoAI (Destination CRM)

• 3CLogic and Glidefast Consulting join forces to transform ServiceNow contact center solutions (Customer Service Manager)

• PyannoteAI raises €8M for speaker intelligence platform (EU-Startups)

• Phonic raises $4 million for speech-to-speech platform (Pulse 2.0)

• GoTranscript embraces AI to evolve jobs, not eliminate them (The Whig)

• VitalPBX 4.5.0 R7 boosts voicemail transcription and AI integration (AiThority)

• Rime introduces Rimecaster for real-time voice streaming (Rime)

• Boson AI introduces HIGGS for audio understanding and expressive speech synthesis (MarkTechPost)

• Google AI composes music, clones voices, and decodes speech (Storyboard18)

• VoicePen transcribes your meetings and lectures into text notes (9to5Mac)

• Modulate unveils VoiceVault anti-fraud voice tech (Modulate)

• Infinix Note 50’s smart AI features now voiced by Davido (Zikoko)

• JobTalent raises €92M in Series F funding (Finsmes)

---

Voice AI Podcast 🎙️

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

Engineering Corner 😎

• Sesame speech model generates human-like speech (Towards Data Science)

• AI can give voice to sign language, empowering the deaf (HackerNoon)

• Voice-first AI projects boost productivity without typing (HackerNoon)

• Speech-to-speech AI: from Dr. Sbaitso to Amazon Nova (Dev.to)

• MultiMedST: New dataset breaks barriers in medical speech translation (Dev.to)

• Behind the curtain: How modern text-to-speech AI works (Dev.to)

• Gemini 2.5 Pro handles 2-hour audio transcriptions (Geeky Gadgets)

• Top AI speech solutions revolutionizing communication in 2025 (CodeCondo)

• Dynamic client selection and group-balanced personalization for data-imbalanced federated speech recognition (MDPI)

• Enhancing far-field speech recognition with Mixer: A novel data augmentation approach (MDPI)

• kNN-SVC: robust zero-shot singing voice conversion with additive synthesis and concatenation smoothness optimization (ResearchGate)

• Best text-to-speech for YouTube videos (Durham Post)