World first: ‘RighValor’ platform delivers distributed agentic AI at the edge for personalised home services
By Claus Hetting, WiFi NOW CEO & Chairman
The broadband world is rushing towards AI everywhere but implementing agentic AI in the home from scratch will likely be pricey and complex at a time when CSPs are under pressure to cut costs. Now startup Righ is proposing a radically different approach: ‘RighValor’ – announced today – is distributed & Wi-Fi connected agentic AI at the edge. The aim is to deliver cost-effective AI-powered services that can even run on legacy HW, the company says.
Broadband service providers will soon be facing an arduous and costly task: Upgrading infrastructures for the purposes of deploying agentic AI. Broadband networks connect many tens of millions of homes and the lure of (possibly) lucrative AI-based home services is no doubt irresistible. At the same time the road ahead is arguably fraught with a litany of technical and financial risks, most of them unknown.
So could there be a better way? CPE and broadband solution provider Righ today answered that question with a resounding yes. Righ’s just released ‘RighValor’ AI and software platform is the company’s bid towards deploying effective agentic AI into homes both faster and cheaper. The new platform is a fundamental rethink of AI service architectures for the home, the company says.
Photo and Quote from Liem Vo, CEO Righ Inc.
Righ says that the new platform ushers in a long list of benefits. First of all, legacy hardware can be used to deliver agentic AI services at the edge. “RighValor has been designed as the AI layer residing atop our multi-protocol RighGravity connectivity framework released earlier this year. This makes the introduction of agentic AI with RighValor really easy, because any CPE or mesh unit or set-top box running RighGravity instantly becomes part of the AI processing network,” Liem Vo says.
The new platform also inherently protects user privacy because inferences and learning takes place within the home, which means no personal data leaves for the Cloud. “We use the Cloud for federated learning, meaning we leverage learnings gained in individual homes towards all homes. With mostly local processing, Cloud costs are minimized and latency on critical tasks is kept low. And the learnings within one home are highly personalized to individual users and homes ” he says.
So what AI-type services can users expect? Personalized Wi-Fi connectivity optimization that caters to the specific devices and usage in a location is one obvious application as is Wi-Fi sensing for occupancy monitoring or security. Sensors for security, heating, air conditioning, lights, and more can be tied together for personalized control of the home while cameras can be used for intelligent personal identification and more. All of such devices within the home are nearly always wirelessly connected, which is why this architecture is possible in the first place.
“You can think of RighValor as the basis for an intelligent, living network and AI service on top of the connectivity framework created by RighGravity. The parts may be small but together they make up the intelligent artificial ‘brain’ for the home. In the future we can imagine personal computers, phones, and perhaps gaming devices as part of the processing pool. And of course as NPUs become available in devices, the processing power will increase,” Liem Vo says.
More about Author, Claus Hetting, CEO of Wi-Fi Now: Wi-Fi NOW is the leadership function at the centre of a global community of tens of thousands of passionate Wi-Fi industry readers and followers. Our readers trust us to bring them the right information on all things Wi-Fi from across the world.
Our mission is to build awareness about the impact of all great Wi-Fi technology out there.
