I didn’t pitch AVOLA. I built it. Because I kept seeing the same gap — brilliant guidance with no mechanism to make it stick. Here is what AVOLA is, why it exists now, and where it is heading.
I have spent 15 years building production AI systems in environments where the stakes are real. At Micron, I deployed AI platforms across semiconductor fabrication facilities — environments where a failed model recommendation can halt a production line and cost millions. At IBM, I built real-time incident management systems where reliability is not a feature request; it is the minimum requirement for the system to exist.
In both environments, I learned the same thing about complex, high-stakes systems: the gap between knowing the right answer and executing the right answer is almost always larger than anyone expects. And that gap is where most value is destroyed.
Over the years, I watched smart, motivated people receive good advice — from doctors, from nutritionists, from wellness programmes — and fail to implement it. Not because they disagreed. Not because they were undisciplined. But because the advice had no execution infrastructure behind it.
A doctor tells a patient with pre-diabetes to reduce refined carbohydrates. The patient agrees. Leaves the clinic. Goes home. Opens the fridge. There is no plan. There are no suitable ingredients. There is, however, a food delivery app with a one-tap reorder button for yesterday's biryani.
The advice was correct. The intention was genuine. The system that would have made the advice executable — the meal plan, the pantry management, the grocery sourcing — simply did not exist.
Preventive healthcare does not fail at the clinic. It fails in the space between the consultation and the kitchen — a space that no product, no service, and no technology had ever seriously tried to fill for ordinary Indian families.
I have been aware of this gap for years. The reason AVOLA exists now, rather than five years ago, is that two things converged in 2024 that made it finally possible to build.
The first is agentic AI. The Claude API and the broader capability of large language models to reason, plan, and execute multi-step tasks means that for the first time, it is possible to build a system that genuinely does what a personal nutritionist does — analyses your health context, builds a personalised plan, adapts that plan based on what you actually have, and executes the grocery sourcing — at ₹399 a month. That was not possible in 2019. It is possible now.
The second is quick commerce. Zepto, Blinkit, and Instamart mean that the gap between “I have a meal plan” and “I have the ingredients to execute it” is now 15 minutes in most Indian metros. The logistics infrastructure that makes AVOLA's execution loop complete did not exist at scale five years ago.
AVOLA is four AI agents working together as a system. NYLA is the brain — she analyses your family's health profiles, generates personalised 7-day meal plans, and delivers weekly nutrition insights. SCOUT sees your pantry through your camera and tracks what you have, what's expiring, and what's missing. CHEF generates recipes from what is actually in your kitchen, filtered by your goals and preferences. HERALD compares Zepto, Blinkit, and Instamart for missing ingredients and opens your cart in one tap.
The system is live in production. 30+ models, 25+ controllers. Subscription paywall, family multi-profile support for up to 6 members, push notifications, adherence tracking — all built and running.
We took this to Brussels in early 2026 for Health-on-Stage — one of Europe's leading health innovation events. The clinical nutrition leadership at UZ Brussels validated the thesis immediately. The gap we are trying to close is not an India-specific problem. It is a global execution gap, and India is simply the most urgent place to address it.
The version of AVOLA that exists today solves the meal planning and grocery execution problem. It is a meaningful intervention — one that, at scale, has the potential to materially reduce the number of Indians who develop preventable metabolic disease between the ages of 35 and 55.
But the longer arc of what we are building is the execution infrastructure for preventive health — not just nutrition, but the full set of daily lifestyle behaviours that determine long-term health outcomes. Sleep. Physical activity. Stress management. These are all, at their core, execution problems. And execution problems yield to systems.
We are a small team. We are early. We are raising capital to validate the model in Bengaluru before scaling. If you believe preventive healthcare should not end with the consultation — if you believe the execution layer is the missing piece — I would like you to be part of what we are building.
Join the waitlist. Reach out. Or just follow along.
— Pradeep Shekaran
Founder & CEO, EpicureAI Labs
Bengaluru, April 2026