Q/A with Pankaj Kedia: Sensors Converge keynoter talks Edge AI

AI thought leader Pankaj Kedia took out time with Fierce to preview his upcoming keynote at Sensors Converge, “The Edge AI Revolution: Next generation devices and sensors.”  Kedia founded 2468 Ventures and now serves as managing partner. Previously he held roles as a senior executive at both Qualcomm and Intel. 

Fierce: What’s the thesis of your upcoming keynote?

Kedia: In my keynote, “The Edge AI Revolution: Next Generation Devices and Sensors,” I will discuss the foundation we are building with Cloud AI, the range of smart devices and sensors we will deploy with Edge AI and the data and intelligence we will perceive, process, and act upon in real-time as we unleash new applications and services with Applied AI. For every dollar of value we create in Cloud AI, we will deliver 10 times the value with Edge AI, and another 10 times with Applied AI.

Fierce: The edge AI revolution is pretty real based on analyst expectations for growth in sales through the next decade. But what is the most obvious evidence of this growth?

Kedia: The most obvious evidence is how we, as an industry, are beginning to take for granted on-device compute with integrated AI Accelerators. These integrated NPUs or AI accelerators are enabling always-on, always-connected intelligent experiences across a range of smart devices — from AI PCs to AI smartphones, from robots to humanoids, from cameras to wearables and from self-driving cars to robotaxis.

Fierce:  We hear reports that pilots in edge AI most often do not result in rollouts. Why is that in your experience?

Kedia: Edge AI represents the next phase of AI investments as we move inference from the cloud to the edge. For Edge AI to be successful, we need to (a) ensure there is adequate compute at low power in these devices, (b) the language models are small enough to fit on the device while being powerful enough to deliver value, (c) there is large volume of labeled real-world data, (d) the resulting user experience preserves the highest degrees of security and privacy, and (e) the new AI systems and devices are seamlessly integrated with legacy systems, sensors, and software.

 In instances where these requirements are duly addressed, I have seen highly successful conversions from pilots to broad rollout. In instances where these are not, the broad rollout is not happy — driven sometimes by technology reasons but many times by organizational readiness reasons.

Fierce: Is it the merging of AI chips with sensors and memory on SoCs that makes this growth possible? What is the driver for most of the technology growth?

Kedia: AI brings intelligence to everyday experiences in its most simplistic definition. While the internet, mobile, and cloud eras primarily focused on disrupting the tech sector, which has grown from sub 5% to over 10% of the global economy over the last couple of decades, the AI era is going to transform not just the tech sector but also the services and labor components, which represent 50% of the global economy.

The primary driver and reason for AI to be bigger than previous tech revolutions is the enormous benefits it brings to almost 60% of the global economy across every sector, in every region, and for every human.

Yes, the convergence of compute, memory, connectivity, and sensors in SOCs that are designed for the AI era is a critical enabler to help us unleash this full promise of AI. From an Edge AI perspective, many of the applications such as autonomous driving or industrial automation require instantaneous response, are built on proprietary datasets, require superior performance/power and performance/cost characteristics, and need the highest levels of security and privacy to scale.

Fierce: What do you want new engineers and others to learn from your keynote?

Kedia: We are at the beginning of the AI and AI device eras, where intelligence will be embedded into everything, everywhere — if it senses, it will integrate smart AI. While in the short-term, we are deploying AI to improve productivity and do things faster, better, cheaper, in the long-term, the biggest breakthroughs will come when we use AI to pursue use cases we have not been tackle in our lives such as curing cancer with drug discovery or eliminating accidents due to driver error with self-driving cars.

I would encourage every attendee to embrace AI and pursue bold transformations in their personal or professional lives.  As we move from Cloud AI to Edge AI and Applied AI, there are unprecedented opportunities for startups and established companies alike to rethink how we work, play, connect, live, and share. What we do over the next decade will determine how we live our lives in this century.

Editor’s Note: Pankaj Kedia will deliver the keynote, “The Edge AI Revolution: Next Generation Devices and Sensors” at 9 am PT on Thursday May 7 at Sensors Converge 2026 at Santa Clara Convention Center in California.. Registration for the May 5-7 event is available online.