Connect with us

Technology

Built in Bharat: Yellow.ai Launches Nexus Vox — the First Enterprise Voice AI That Can Clone Any Voice and Deploy It Across 500+ Languages in Under a Second

Published

on

Yellow.ai turns 10, and the Bengaluru-born company has launched the first enterprise voice AI forged by a decade of experience in solving India’s hardest conversational AI problems. Vox can deploy across 20+ Indian languages, and 500+ languages globally.

BENGALURU, India, May 6, 2026 /PRNewswire/ — Ten years ago, three friends sat in a room in Koramangala with an idea: that businesses could talk to their customers through AI — in the languages they actually speak. India’s startup scene was chasing food delivery. Enterprise AI wasn’t a category.

A decade later, Yellow.ai has powered billions of conversations for 600+ enterprises across 85 countries. And in its 10th year, the company is launching the product that a complex and demanding market like India needs.

Today, Yellow.ai announced Nexus Vox — the first enterprise voice AI built as a single integrated system, not stitched together from multiple vendors’ APIs. The product supports 500+ languages and dialects natively — including all major Indian languages, plus Hinglish and dozens of regional dialects. With end-to-end latency below 400 milliseconds which is within the threshold of natural human conversation.

For Indian enterprises, this changes everything.

The Problem India Knows Better Than Anyone

India runs on voice. A nation of 1.4 billion people, hundreds of languages and dialects. Enterprise voice automation has never been able to keep up.

The reason is architectural. Today’s enterprise voice systems are Frankenstein stacks — one vendor for speech recognition, another for voice synthesis, a third for conversational AI, a fourth for telephony. Every handoff adds 100–200 milliseconds of latency. Round-trip response times hit 800 milliseconds or more — long enough to make every conversation feel robotic. And most platforms support fewer than 30 languages, leaving the vast majority of India’s population on hold or hanging up.

“Enterprise voice AI has been a Frankenstein’s monster — several vendor APIs stitched together, each adding latency, each introducing a point of failure, and none of them effectively resolving the customer’s problem,” said Raghu Ravinutala, Co-founder and CEO of Yellow.ai. “We’ve spent 10 years solving India’s hardest enterprise AI problems. India taught us that if your AI doesn’t speak the customer’s language — fluently, instantly, in the accent they think in — it’s not AI. It’s an IVR with a better UI. Vox is built with a completely different architecture. “

How Nexus Vox Works

Nexus Vox runs natively inside Yellow.ai’s Nexus platform, sharing the same runtime as the conversational AI, knowledge base, and automation engine, what Yellow.ai calls a “zero-hop architecture.” There are no API round-trips between voice processing and conversation processing.

Key capabilities:

Sub-400ms end-to-end latency from customer speech to AI voice response — within the range of natural human conversation.500+ languages and dialects with natural, human-quality voice synthesis10-second voice cloning from any audio sample — preserving the original speaker’s timbre, cadence, and emotional range. An Indian enterprise can clone its own brand voice and deploy it across every language it serves.Real-time sentiment awareness — Vox adjusts tone, pacing, and escalation behaviour mid-conversation, critical for high-emotion calls like insurance claims and service complaints.Omnichannel deployment across telephony (SIP/PSTN), web widgets, and REST APIs from a single configuration.CX and EX on one platform — built for both customer-facing and employee-facing voice automation. Not two separate products.Sarvam AI integration — Vox is natively integrated with Sarvam AI’s sovereign Indian language models, built under the Government of India’s IndiaAI Mission and trained from scratch on Indian language data.

Forged by India. Ready for the World.

Yellow.ai’s ability to build Nexus Vox is inseparable from its decade in India.

Edelweiss General Insurance launched India’s first end-to-end AI Voice Bot for motor claims — powered by Yellow.ai, processing Hinglish from noisy garages, 24/7. A claimant calls after an accident, stressed, switching between Hindi and English mid-sentence, and the AI handles the entire intimation — names, addresses, licence numbers — in real time. That’s the kind of problem that forces you to build voice AI differently.

Sony India deployed Isha — a voice AI agent providing round-the-clock support in English, Hindi, and Bengali, handling 21,000+ customer and dealer calls within two months of launch.

Use cases like these pushed Yellow.ai to solve problems no other market demands at this scale and linguistic complexity. Nexus Vox is the product those problems made inevitable.

The business reflects the bet. Yellow.ai is now EBITDA positive — a rarity in an industry where most AI companies are still burning cash. Revenue has grown to multiple 10s of million dollars at a CAGR of 70%, with 40% consolidated growth globally. And rather than pulling back, the company is increasing its investment in India, expanding its Bengaluru engineering centre, deepening its partnerships in the region.

“Every AI company talks about India as a future market. For us, India has been the market from day one — the one that made us profitable, the one that trained our platform, and the one we’re investing deeper into,” said Ravinutala. “India is the hardest AI market on earth — hundreds of languages, a billion WhatsApp users, enterprise buyers who demand world-class AI at Indian economics. We don’t see that as a challenge. We see it as the reason we were able to build Vox. India didn’t just birth Yellow.ai. India trained us to solve for scale.”

What Nexus Vox Means for Indian Enterprises

For India’s largest industries, Nexus Vox unlocks voice automation that was previously impossible:

Banking & Financial Services: A large Indian bank can now deploy a single AI voice agent that handles customer calls in Hindi, Tamil, Telugu, Bengali, Marathi, Gujarati, and English — without separate vendors per language. First-call resolution improves. The customer in Lucknow and the customer in Chennai get the same experience, in their own language, at any hour.

Insurance Claims Intimation — the most voice-dependent, high-emotion process in Indian insurance — can now be fully automated in Hinglish and regional languages, with real-time sentiment awareness that detects distress and escalates appropriately.

Airlines & Travel: An airline handling 100+ million passengers can automate rebookings, cancellations, and loyalty queries on voice in every Indian language — not just English and Hindi — across phone and WhatsApp from a single deployment.

Public Sector & Citizen Services: Government agencies and public enterprises can deploy voice AI that serves citizens in their mother tongue — from LPG bookings to tax queries to healthcare information — without language becoming a barrier to access.

Availability

Nexus Vox is available immediately for enterprise customers on the Yellow.ai platform. Indian enterprises can request a personalised demo, including a live voice clone demonstration in any Indian language, at yellow.ai/nexus-vox.

About Yellow.ai

Yellow.ai is the global leader in enterprise conversational and generative AI, trusted by 600+ enterprises across six global markets including Waste Connections, Ferrellgas, Randstad, Tiket.com, and Hyundai. The company’s Nexus platform — the industry’s first Universal Agentic Interface — powers autonomous customer and employee experiences across voice, chat, and email channels. Yellow.ai is headquartered in San Mateo, California, with offices in Bangalore, Jakarta, Dubai, and London. For more information, visit yellow.ai.

Photo: https://mma.prnewswire.com/media/2972716/Nexus_VOX.jpg
Logo: https://mma.prnewswire.com/media/2061481/5953647/Yellow_ai_Logo_Logo.jpg

 

View original content to download multimedia:https://www.prnewswire.com/in/news-releases/built-in-bharat-yellowai-launches-nexus-vox–the-first-enterprise-voice-ai-that-can-clone-any-voice-and-deploy-it-across-500-languages-in-under-a-second-302762752.html

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Technology

He Sold Everything and Moved to Colombia for Her. Then Built the Translator They Needed and Named It After Her.

Published

on

By

Nayerly, a patent-pending iPhone app, is the first translator both people can talk over at once. Fully offline, 22 languages.

EDMONDS, Wash., Aug. 25, 2026 /PRNewswire/ — Michael Wilson sold everything he owned and left the US for Colombia to be with the woman he loved. They did not share a language.

Every translator they tried forced them to take turns: one speaks, stops, waits for playback, and only then can the other answer. It worked. It also turned every conversation into a transaction.

So he built his own, alone, and named it after her. Nayerly is now on the App Store.

Nayerly is full duplex, meaning it listens and speaks at the same time instead of taking turns. The app keeps listening while it is speaking, so you can talk over the translation, and it keeps listening when more than one person talks, so two people can talk over each other and it still works. No phone passed back and forth. It works like a personal interpreter sitting between you.

“Every other translator is a walkie-talkie. You talk, you stop, you wait,” Wilson said. “I built this so I could speak better with someone I love. That is the only way you make a real connection. It works so well you forget you are using it and start talking over each other.”

Everything runs on the device. All 22 language models download during setup, so it works with no Wi-Fi and no signal – on a plane, abroad, off the grid – and no audio ever leaves the phone. Hands-free works with any AirPods, not just the premium ones.

Key features:

– Full duplex: listens and speaks at the same time, so you can talk over the translation and over each other
– 100% offline in 22 languages, including Spanish, Arabic, Chinese, Japanese, Hindi and Ukrainian
– Hands-free with any AirPods
– Private by design: audio is never recorded or stored
– Free to download, 10 minutes of translation daily

Unlimited use is $1.99 per month during launch, billed annually. A one-time lifetime option includes Apple Family Sharing. New languages will be added in future updates at no additional cost to existing users.

Videos of Wilson and his girlfriend using Nayerly in Colombia, both speaking at once in English and Spanish, are on Instagram at instagram.com/nayerlyapp and TikTok at tiktok.com/@nayerlyapp.

Availability

On the App Store for iPhone now: https://apps.apple.com/app/id6762064110
More: https://nayerly.com 

About Nayerly
Nayerly is an independent iPhone app built alone by Michael Wilson and named for the woman he built it for. It exists so two people without a shared language can just talk.

Media Contact
Michael Wilson
Founder, Nayerly
421311@email4pr.com
425-220-5237

View original content to download multimedia:https://www.prnewswire.com/news-releases/he-sold-everything-and-moved-to-colombia-for-her-then-built-the-translator-they-needed-and-named-it-after-her-302859030.html

SOURCE Nayerly

Continue Reading

Technology

Eleven GovTech Startups Named to Pitch at State of GovTech 2026

Published

on

By

Companies from across the country will pitch live to government and investor judges in Montgomery County, Maryland, September 9–10. Registration is free for government employees.

SILVER SPRING, Md., Aug. 25, 2026 /PRNewswire/ — CivStart today announced the eleven startups selected to pitch at State of GovTech 2026, the organization’s seventh annual national govtech summit, taking place September 9–10 at the Silver Spring Civic Building in Montgomery County, Maryland.

The selected companies span permitting, public safety, emergency management, grant administration, constituent services, and civic infrastructure finance. They will pitch live on the main stage to a judging panel drawn from local government, county technology leadership, and the govtech investment community.

The 2026 Startup Cohort

Appellate  (Bentonville, AR) — system of record for police towing and vehicle custodyAyla  (Washington, D.C.) — public-sector recruiting on a live government labor-market data engineCODICE  (Washington, D.C.) — no-code permitting, licensing, and inspections platformGostly  (Westminster, MD) — real-time command center for urban infrastructureGrantwell  (Los Angeles, CA) — AI grant management across the full lifecycleGreen Stream Technologies (Wake Forest, NC) — flood and weather early warning systemsGovstream  (Seattle, WA) — AI-driven permitting intakeLabrynth  (US, Australia, EU, UK) — AI-native regulatory intelligence for permittingManyBond — crowdfinancing for community-supported local investmentReadyly  (Connecticut) — agentic AI resolving resident inquiries across every channelSamora AI  (Bellevue, WA) — multilingual voice AI for citizen information lines

The Judging Panel

Hillary Orr, Deputy Director of Transportation, City of Alexandria, Virginia Rita Reynolds, Director of Public Sector, CAI Thao Hill, Co-Founder and CEO, Govinity Tom Spenger, CEO, SOVRA Eyal Feder-Levy, CEO, Zencity

The pitch competition is one part of a two-day agenda built around working sessions rather than passive programming. Attendees draft AI use policies, run live cybersecurity breach simulations, and bring unresolved problems from their own jurisdictions to work through with peers.

State of GovTech 2026 is expected to draw more than 200 attendees, split roughly evenly between public sector leaders and the entrepreneurs, investors, and industry practitioners who serve them. The venue sits directly on the Washington Metro, and registration is free for government employees. Full agenda and registration at civstart.com/sogt26.

About CivStart

CivStart is a govtech organization connecting government leaders with the technology solutions and entrepreneurs solving public sector challenges. Through its accelerator programs, industry events, and products including Clarity and GovFit, CivStart helps state and local governments define their challenges and find partners equipped to solve them. State of GovTech is CivStart’s flagship annual summit, now in its seventh year. Learn more at civstart.com.

Media Contact:

Nick Lyell

Co-Founder & COO

421351@email4pr.com | 608.234.2166

View original content to download multimedia:https://www.prnewswire.com/news-releases/eleven-govtech-startups-named-to-pitch-at-state-of-govtech-2026-302859026.html

SOURCE CivStart

Continue Reading

Technology

Goodix Unveils New Generation Touch Screen Controllers for Medium and Large Displays

Published

on

By

SHENZHEN, China, Aug. 25, 2026 /PRNewswire/ — Goodix has launched its new-generation GT9976 series of high-performance touch screen controllers for medium and large displays, targeting premium flexible OLED tablets, foldable smartphone main screens, and other smart devices. Delivering major upgrades in gaming control, touch performance in challenging environments, and active stylus support, the series enables device makers to create differentiated large-screen products for gaming, work, and learning.

As an early commercial deployment of the series, Lenovo’s latest Legion Y700 Wuji gaming flagship tablet and Tab Pen Pro 2 integrate Goodix’s innovative portfolio, including the GT9976N touch screen controller, an active stylus driver chip, and an ultra-narrow side-key capacitive fingerprint sensor, to deliver a more responsive and immersive user experience.

Faster Response for Smoother Large-Screen Interaction

High-refresh-rate displays used in competitive gaming place greater demands on touch response and data processing. The GT9976 series features Goodix’s proprietary full-screen parallel sensing architecture, which achieves a higher signal-to-noise ratio (SNR) within a shorter scan time — significantly accelerating finger-signal capture and processing efficiency.

Supporting 7- to 9.5-inch flexible OLED displays, the new series delivers industry-leading touch report rates and instant sampling rates. From everyday browsing and app switching to sudden stops, sharp turns and rapid multi-taps in high-frame-rate games, the GT9976 series ensures smooth, precise, and highly responsive touch interaction.

Reliable, High-Precision Touch Performance in Demanding Conditions

To address the noise interference from flexible OLED panels under heavy system loads, the GT9976 series adopts an industry-leading high-noise-suppression analog front end (AFE) sampling architecture that doubles SNR over the previous generation. Combined with Goodix’s proprietary gaming algorithms, the controllers accurately identify valid touch signals even in complex noise environments, delivering stable, precise coordinate data at all times.

In demanding scenarios such as FPS, MOBA and rhythm games, the GT9976 series combines a higher SNR, high-speed scanning and sampling, and enhanced real-time noise monitoring with frequency-hopping technology to enable faster tap response and stable multi-finger control without jitter or drift. This improves responsiveness and tracking accuracy for pixel-level actions such as flick aiming, recoil control, combo execution and movement.

The series also integrates advanced dual-mode self- and mutual-capacitance sensing with complementary algorithms to improve palm rejection, effectively reducing false touches caused by grip and edge contact. Dedicated optimization for sweaty hands, wet fingers and gaming finger sleeves helps maintain sensitive and stable touch performance under challenging real-world conditions.

Enhanced Active Stylus Support for Productivity and Creativity

For mobile productivity, education and professional drawing, the GT9976 series also delivers significantly enhanced active stylus performance. When paired with Goodix active stylus driver chip, the solution supports a 360Hz pen report rate. Together with Goodix’s proprietary stylus anti-interference algorithm, it achieves outstanding linearity and precision — accurately reproducing every stroke for tasks such as note-taking, fine annotation, sketching, and handwriting practice.

For flagship OLED tablets with large screens, Goodix also offers the GT7A series, which supports display sizes ranging from 10 to 14 inches while delivering the same high overall performance, further strengthening its portfolio for mid- to large-size touch applications. Both the GT9976 and GT7A series have been validated by leading device brands and display panel manufacturers, and are set to be adopted in foldable smartphones, gaming tablets, and slim flagship tablets — enabling smoother, more precise, and more reliable large-screen interaction.

View original content to download multimedia:https://www.prnewswire.com/news-releases/goodix-unveils-new-generation-touch-screen-controllers-for-medium-and-large-displays-302859395.html

SOURCE Goodix

Continue Reading

Trending