xAI Ships Voice Agent Builder, Krisp named 2026 Disruptive Technology
Voice AI weekly digest
Events
Real-Time Observability for Production Voice AI Regal walks through its Observability Dashboard with real catch-and-fix examples from production voice agents. (Jul 9, SF, Voice AI Space)
AI Tinkerers San Francisco: July GTM Engineering Track w/ Attio Builder-only, no-pitch meetup with live GTM engineering demos - a solid room for voice agent developers. (Jul 8, SF, AI Tinkerers)
Top Updates 💪
xAI launches Voice Agent Builder, a no-code platform for production voice agents with telephony and guardrails at $0.05/min. (xAI)
Krisp named 2026 Disruptive Technology of the Year by CMP Research for its voice AI infrastructure. (Krisp Blog)
ElevenLabs explores a $22B tender offer, doubling its valuation from the $11B Series D five months ago. (Tech in Asia)
Pocket raises $11M from Accel and YC for its $129 AI note-taking puck that has shipped 130K units. (TechCrunch)
Lucida AI raises €6.1M seed for its speech-to-speech language coaching platform, now at 3M users. (EU-Startups)
US senators revive the AI Labeling Act, a bipartisan bill requiring AI-generated audio and video to carry disclosure labels. (Music Business Worldwide)
Retell AI launches Conductor, a graph-native review interface with an AI copilot for production voice agents. (AIthority)
Syntiant and Vibe partner to bring voice-enabled AI to smart workspace hardware using edge AI chips. (GlobeNewsWire)
HealthLynked launches an AI healthcare platform with 24/7 scheduling and medical office voice agents. (GlobeNewsWire)
RevComm launches MiiTel for Retail, extending its voice AI analytics to in-store customer conversations. (IBTimes JP)
Patient trust is the biggest barrier to healthcare voice AI, not the technology, argues a Forbes analysis. (Forbes)
Voices similar to our own are more persuasive, finds new research raising concerns about companies weaponizing stored voice data. (Nautilus)
Synthflow deployed voice AI in one day for Nellis Auction, now handling 80% of each customer interaction automatically. (CXM World)
Engineering Corner 😎
Mozilla AI releases transcribe-cpp, an open-source C/C++ STT library with ggml runtime and GPU acceleration. (StartupHub)
ViiTorVoice-NAR goes open source with word-level TTS editing that swaps individual words without regenerating surrounding audio. (TechTimes)
Higgs TTS 2 3B from BosonAI is a 5.8B-param TTS model trained on 10M+ hours with zero-shot voice cloning. (HackerNoon)
Vowen 0.4.8 released, a free offline voice productivity app using Whisper-based local transcription. (Warp2Search)
WhisTam is a Whisper-based framework for Tamil dialect speech recognition, ranking 2nd at DravidianLangTech@ACL 2026. (ACL Anthology)