artash.lens .ink (✸,✸) 🦭/acc 🍚base.eth 🪞Noya

@dtelecom If your agent hallucinates speech during silence or reacts too slowly when a user interrupts, the issue often starts in the same place: the audio pipeline.

@bbcpersian بالاخره یک تحلیلگر درست و حسابی توی این بی بی سی دیدیم

دونالد ترامپ بار دیگر تهدید کرده که اگر تهران به توافق تن ندهد، تمام نیروگاه‌های برق این کشور را منهدم و ایران را به «عصر حجر» بر می‌گرداند. بعد از پایان صحبت‌های او اسرائیل اعلام کرد این کشور هدف موشک‌های ایرانی قرار گرفته. شایان سمیعی، تحلیلگر روابط بین‌الملل:

@dtelecom Voice AI is moving into a much larger market cycle: GVR projects the voice and speech recognition market to reach $53.67B by 2030.

@BinanceWallet So bullish on $check

Checkmate (CHECK) is now live on Binance Alpha! Users with at least 242 Binance Alpha Points can claim an airdrop of 500 CHECK tokens on a first-come, first-served basis. If the reward pool is not fully distributed, the score threshold will automatically decrease by 5 points every 5 minutes. Please note that claiming the airdrop will consume 15 Binance Alpha Points. Users must confirm their claim on the Alpha Events page within 24 hours; otherwise, it will be deemed that users have given up claiming the airdrop.

@dtelecom where Agents and Human talk, listen, interact.they have created a new communication layer. Decentralized. Real-time. AI-native.

@Americanfort_io Cool

Anyone can Google your wallet and see everything such as balance, history, every move. We fixed that. One name. Any chain. Zero exposure. No mixers. Fully compliant. Quantum-resistant. Join the airdrop & earn rewards quests.americanfortress.io

@CoinbaseMarkets $check mate

Spot trading for Checkmate (CHECK) and Sign (SIGN) will go live on 2 April 2026. The opening of our CHECK-USD and SIGN-USD trading pairs will begin on or after 9AM PT, if liquidity conditions are met, in regions where trading is supported.

@dtelecom a better approach is to break that work into stages. dTelecom applies VAD first, then denoising, then speech validation, then transcription, followed by hallucination filtering after decoding.

@dtelecom If a team needs a broader voice catalog, faster deployment, less infrastructure ownership, or a fully managed path to scale, API-based TTS still has clear advantages

@dtelecom LFG

One center. One owner. One point of failure. That era is over. We’ve created a new communication layer. Decentralized. Real-time. AI-native. @dTelecom — where Agents and Human talk, listen, interact.

The same direction described in @dtelecom recent STT write-up and in Whisper community discussions around phantom subtitles, repeated phrases, and hallucinated lines produced during silence.

@dtelecom The first few seconds of a call usually show whether a real-time voice agent is built well. Two bugs usually decide the outcome.

@konnex_world @JOllwerther New era is coming

"The era of isolated machines is over." Meet our CEO, @JOllwerther . Business leader with 15+ years in robotics, drone-tech, and media. Jon led commercial teams at Measure (Drone as a Service) and Aerobo/Vermeer. Now, he’s creating TCP/IP for Motion. Get ready to experience Konnex.

@czrexchange good

All big crypto exchanges started before the crowds got there. Before that, there was a window to be early. It's happening again. This time, you're here. Take action & reap the rewards. quest.czrex.com

@dtelecom That TTS delay is one of the most noticeable parts of the loop. Shaving ms off the speech stage made the interaction feel faster, tighter, and more natural.

@dtelecom describes this kind of multi-stage path in production, and the Whisper ecosystem points to the same pattern through VAD-based preprocessing and timestamp offset correction.

@dtelecom Good

@dtelecom Decentralized 1v1 calls are here, create your account, find ppl, team members by usernames and bio, fav your closest contacts