khaa-lo built AI search marketing intelligence for emerging brands. Its next move is understanding the prompts.

The tools for winning AI search were built for companies that were already winning. Vaishnavi Varma is betting the more valuable signal runs the other way, from what shoppers ask back to the brands to act on it.

TL;DR

khaa-lo began by helping brands navigate AI search. Founder Vaishnavi Varma is expanding it into a marketing intelligence platform that helps brands understand consumer intent, emerging search behavior, and market trends while connecting consumers with relevant emerging brands.

When a shopper asks an AI assistant which protein powder suits a sensitive gut, or which small-batch skincare line will not wreck oily skin, they hand over something a search keyword never carried. The phrasing is specific, frequently clumsy, and it arrives well ahead of any purchase or trend report.

Most of that signal goes nowhere useful. khaa-lo is built to catch it. Founded by AI product designer Vaishnavi Varma, the platform started as AI visibility tooling for emerging consumer brands and is expanding into a discovery product that understands consumer intent directly from AI search.

What it shipped first

khaa-lo launched as AI discoverability software for small consumer labels, built on a straightforward observation about timing. When Varma started, most services addressing large language model discovery were aimed at enterprises. Her concern was that young brands failing to adapt early would be quietly sorted into obscurity by systems they never learned to read.

khaa-lo began with a simple observation: AI search was changing how consumers discover brands, but most companies still measured demand through traditional search, social media, and historical market research. Varma’s early work focused on helping consumer brands understand how they appeared in AI-generated answers, an area that was still largely overlooked outside enterprise software.

As the market evolved, so did the company. Rather than focusing solely on discoverability, khaa-lo expanded into a marketing intelligence platform designed to help brands understand the signals emerging from AI search itself. The product recognizes patterns in consumer intent, evolving search behavior, and market trends, giving marketing teams a new source of intelligence for product development, positioning, and growth.

The expansion

The second phase inverts the direction of the information. Varma is extending khaa-lo into an indie CPG discovery platform that works both sides of the same exchange: brands understand what consumers are asking AI search in their category, and consumers get a route to small and emerging labels that fit what they say they are trying to do. She spent more than two years working with consumers and consumer businesses before building it, specifically to understand how AI search was changing what people buy.

Her critique of the existing tooling market is that it solves the wrong end of the problem. Most vendors help businesses construct a narrative engineered to become the answer to a given prompt. Companies with existing visibility, content volume and domain authority are already legible to AI systems, so they benefit first and most.

The second-order effect is the one she is more interested in. Optimising for what a brand thinks a shopper wants widens the distance between the two. khaa-lo is built to let brands read the demand directly.

The market data supports the opening she is describing. Research presented at Adobe Summit this year, drawn from a database of more than 213 million language model prompts, found that 62% of brands are technically invisible to generative AI models, and that only 8 to 12% of the results appearing in AI answers overlap with those ranking well in traditional search. The authority emerging brands could never afford to build in search does not transfer into the systems replacing it, which levels a field that was never level.

What that levelling has not touched is access to the tooling. The research came from Semrush, which Adobe absorbed in April in a $1.9 billion deal. Much of the tooling built to interpret this shift has been designed around the needs of enterprise marketing organizations, leaving a gap for growing brands navigating AI-native consumer discovery. khaa-lo’s argument is that the competitive advantage will belong to brands that can interpret consumer intent in AI search before it becomes conventional market research. Hence, AI search should become a strategic capability for growing brands, according to Varma.

Logging without prescribing

The consumer side carries the bigger design risk. According to the company, in the next feature launch, users will be able to upload receipts and track aspects of daily wellbeing that fall outside the usual step counts and heart rate readings, including areas such as protein intake and gut health.

The company is explicit that this is not a health app. It offers no diagnosis and no medical advice. The purpose is reflection and logging: noticing what you actually consume, building a record of it, and using that record to find smaller brands worth trying.

Holding that line is harder than stating it. Consumer wellness products drift toward prescriptive language because prescription converts, and the gap between “here is what you logged” and “here is what you should do” closes quickly under growth pressure. Varma has designed against that pressure deliberately, which places khaa-lo in a quieter corner of the wellness market, one organised around conscious consumption and taste instead of clinical authority.

The receipt upload is the sharper mechanic. Purchase history is among the least performative records a person keeps. It captures what someone bought rather than what they told a feed they aspired to buy, which makes it a better input for brand discovery than engagement data.

Physics, audits, and a UX certification

Varma’s route into product design ran through enterprise security. She holds a physics degree from Syracuse University and moved into vulnerability management at Bank of America, working with large and complex datasets and building dashboards that had to satisfy audit requirements while staying legible to cross-functional leadership.

Those two demands pull against each other constantly. Dashboards built to satisfy an auditor tend to bury the person reading them, and the ones that read easily tend to leave out what the audit requires. She took a Google UX design certification to close that gap, her first formal step into the discipline.

She now runs that thinking as a consulting practice for founders at zero to one, the stage where a product works and no stranger can tell why it matters. The vulnerability management background shows up as a discipline layered on top of the speed. Building fast means the questions that used to get argued over during a slow build now get skipped: how data moves through the system, and where the security model has holes nobody has gone looking for. She wants security and user trust settled in the design process before the product hardens around them, at whatever pace the team is working.

The work itself is unglamorous and mostly upstream of code. It covers which screens a stranger sees first and what a non-technical founder can still maintain once the consultant leaves. It also covers the question most AI products skip, which is whether a feature should run slower than it technically can so the person using it understands what just happened. Founders arrive asking for an interface and leave having had the harder conversation about what their product is for.

“Technology to aid you not replace you” is her motto. It reads as modest until you notice how much of the AI product market is built on the opposite premise.

The art is the method

The strand of Varma’s work that sits furthest from software is the one she considers central. She works in mixed media, using canvas and electric paint, then carries those textures and narratives into functional websites built to communicate clearly to both people and AI search systems. She hand-draws the UI and UX for a founder’s app before opening a design tool. Her demonstrations extend into AR art and interactive installations, exploring how creative expression can become a layer of product experience and discovery.

Underneath the practice is a claim about what small brands lose when they optimise. A founder-led label has a cultural voice that gets sanded down in the process of becoming machine-legible, and Varma treats the sanding as a design failure rather than a necessary cost. Art, for her, is one of the strongest forms of human connection available when no human is there to make it.

That instinct is also what she is selling. The founders who come to her tend to arrive with something that works and no way to show it to anyone, and the fix is rarely more technology. It is someone willing to sit with the product until it explains itself.

What she is building toward

The category Varma is working in has a known failure mode. Plenty of tools promise to show brands how machines perceive them, and most resolve into charts that describe a problem without helping anyone act on it. The same pressure is arriving across consumer AI, where the business model tends to decide what the product eventually becomes.

Varma has spent her career on the opposite problem. The vulnerability dashboards at Bank of America had to make risk legible to executives who did not have time to decode it, and the same requirement runs through everything she has built since: marketing intelligence that helps brands recognize emerging patterns in consumer intent, and a consumer product meant to carry a person’s own description of what they need all the way to a brand without flattening it into a metric. Five years across cybersecurity, fintech, health tech and consumer technology have gone into one design problem at escalating scale.

That is the bet underneath khaa-lo. As AI search becomes the layer through which people find what to buy, the advantage should belong to companies that keep the human legible on both sides of the exchange, and Varma is designing for a version of that shift the rest of the market has largely priced out of reach.