Introducing IrisGo’s New Head of Product, Nagesh Pobbathi

When we founded IrisGo, our core mission was clear: turn raw AI capabilities into an active, on-device assistant that actually executes workflows right alongside you. Having spent over 12 years at Apple leading teams on Siri, I learned that ground-breaking AI technology is only as good as the user experience surrounding it.

Now, as IrisGo expands across major OEM partnerships and deepens its local context engine, we are moving into a critical phase: turning complex local AI architectures into an indispensable desktop experience. To lead this evolution, we needed a Head of Product who understands scale, deeply understands developer tools and platform engineering, and knows how to build products users genuinely love.

That is why I’m thrilled to welcome Nagesh Pobbathi to the team. Nagesh brings 15+ years of experience scaling platforms: securing 100M+ accounts at Capital One, driving $300M in revenue at Yahoo!, and founding Quest AI, a design-to-code platform that served 200K developers. He holds 6 patents in search and personalization.

I sat down with Nagesh to get his perspective on where IrisGo is heading:

Why IrisGo & Why Now?

Jeffrey: You’ve built startups and scaled enterprise platforms at Capital One. What was the main spark or ‘aha’ moment that drew you to IrisGo?

Nagesh: I believe the next step in the evolution of AI is moving from the current “pull” model, where users go to an app and interact with AI, to a world where AI is built into every device, works organically across apps, and is fully aware of the user’s context, needs and preferences.

That could truly be a superpower for anyone who wants to do more. Imagine a small business owner whose most tedious work like invoices, vendor follow-ups, and scheduling gets done automatically exactly the way they’d do it themselves.

As I was exploring building such a system myself, I came across IrisGo with its vision to be the AI Operating System (AI OS) for personal devices. You, Arnold Yip, and Lman Chu have built great products at Apple and elsewhere, the company is backed by Andrew Ng’s AI Fund, and the company's OEM distribution strategy puts it in front of millions of users. So joining forces was an easy call.

On-Device AI vs. The Cloud

Jeffrey: How does an on-device AI OS change the desktop experience compared to standard cloud chatbots?

Nagesh: On-device AI provides an always-on, context-rich experience that is fast, secure, and private. But the state-of-the-art models are still too big to run on devices today, so cloud models give the best experience for hard tasks.

Two things are shifting the balance towards on-device: First, as hardware capabilities of personal devices accelerate, we’ll see more “good enough” models run locally. Second, I believe we will transition from the current general-purpose frontier models to a mixture-of-models approach, where the right/specialized model is called for the task at hand.

So I think the winning experience is hybrid: on-device for everything personal and always-on, cloud for the hardest tasks. You keep sensitive data private on device, and tap into the power of the cloud only when a task needs it. The product challenge is making that routing feel seamless while giving users clear control over what stays on device and what can go to the cloud.

Top Priorities: Head of Product

Jeffrey: As we partner with major PC manufacturers, what are your top product priorities for scaling IrisGo?

Nagesh:

• ●Simplicity. Make it dead-simple for everyone to use AI, right out of the box with zero setup, and works across device configurations and form factors, from PCs to handhelds and more.
• ●Daily habit. Assist with and automate tasks that people do every day and important tasks that matter the most for work and life. That means truly understanding user context with agents that learn over time.
• ●Earned trust. Give users full control over permissions and access, keep user data (shared and inferred) on the device. Build security into the agent and every tool it touches. Keep the user in the loop by default, and make every action predictable and reversible.

Fun Facts

Jeffrey: What’s your favorite desktop app or productivity tool that you can’t live without?

Nagesh: Claude Code (as a plugin in my IDE). While I started off using it as a coding tool, it has now become my go-to for everything coding or not.

The Notes app is also a staple and has become the repository of my life, everything from travel plans, shopping lists, to-dos, to scanned copies of kids homeworks or other documents. Between Claude memory, Notes, and Photos, about 90% of my life is documented there.

And IrisGo’s Radar is quickly becoming my go-to organizer - it automatically tracks and aggregates my commitments across email, calendar, slack, etc. And the agent helps me finish things that I tend to usually postpone. For example, it recently helped me finish a home project by looking up and finding the parts needed on homedepot.com, adding them to the cart, and helping me with step by step visual instructions on how to build it!

Jeffrey: Outside of building products, what do you enjoy doing to recharge?

Nagesh: Traveling with my family, mostly outdoors and nature trips since we have young kids. We recently visited Machu Picchu and hiked the Inca Trail, loved every minute of it.

I started learning to play bansuri, the classical Indian (hindustani) flute about two years ago. Listening to music has always been my refuge, and I am excited I can finally play it.

About IrisGo

IrisGo is the AI Operating System (AI OS) for AI PCs, delivering a contextually-aware, intelligent workflow layer that learns from user behavior to automate tasks across applications. Backed by Andrew Ng’s AI Fund and created by the founding engineers behind Apple's Siri and hardware industry experts, IrisGo is focused on making software that's more adaptive, context-aware, and capable of working on behalf of users. For more information, visit irisgo.ai.