The Marketing Kernal co-founder, an engineer, screenwriter and novelist turned software founder, explains why international businesses are underprepared for AI search, and why he refuses to guarantee results in a field full of promises.
Vaibhav Palhade’s career does not follow a straight line. Trained as a computer engineer in India, he started out as an assistant writer and assistant director, wrote screenplays and dialogue for Hindi films, series and Marathi daily soaps, won awards for films on social themes, published several novels including Echoes of Yesterday in 2024, and spent years in search and digital marketing for software companies selling overseas.
In 2026, after roughly two to three years of research with technology co-founder Pratyush Shrivastava, he co-founded Marketing Kernal, which works on AI visibility: how businesses are represented when customers ask tools such as ChatGPT, Gemini and Perplexity for recommendations. The company is also developing a family of software products, starting with Outreach Kernal, a B2B prospecting tool due in beta in November.
This interview has been edited for length and clarity.
Start with the problem. What is actually changing?
Palhade: Where the decision gets made. For twenty years, search gave people a list and they clicked through to decide. Now an assistant often decides on their behalf and explains why in a paragraph. Pew Research looked at Google results last year and found the median AI summary was 67 words. If you are a business, that is your shop window now. You might get one sentence in it. You might get nothing.
Is that a real problem or a marketing problem?
Palhade: It is a revenue problem that shows up as a marketing problem. The same Pew study found people clicked a traditional result in 8% of visits when there was an AI summary, against 15% when there was not. If fewer people click, the description inside the answer matters much more. And most companies have never read theirs.
How is this different from search engine optimization?
Palhade: SEO is still the foundation. Anyone who says otherwise is selling something. The difference is that a model needs to recognize you as a clear entity, find consistent descriptions of you from credible sources, and be able to read your information without guessing. A page can rank well and the company behind it can still be vague, contradictory or invisible to a model.
You say international businesses are especially exposed. Why?
Palhade: Because there is no single AI answer. A company can be well described in English and almost absent in Arabic or French. It can be recommended in the United States and ignored in the UAE. Regulation also differs. Europe has GDPR and the EU AI Act, and Quebec has Bill 96 for French-language commerce. A global brand can have a dozen different AI reputations and not know about any of them.
You wrote for daily soaps and films before this. How does a screenwriter end up here?
Palhade: More directly than it sounds. A daily soap teaches you that an audience decides in seconds whether to stay, and that you have to earn them again tomorrow. Fiction teaches you that every sentence has to earn its place. An AI answer is the most compressed version of both that I have seen. The question I care about is whether that one sentence about a company is accurate, and whether the company had any say in it.
You have been open about earlier ventures that failed. What did they teach you?
Palhade: Three or four of mine did not survive, and Pratyush had his own startup close. The lessons were ordinary. We built before we validated. We chased volume. We lost patience. So this time we spent years researching before we incorporated. It was frustrating at times. It is also why I think this attempt is better prepared.
Where does India fit in? Many of your peers would start with the US alone.
Palhade: India is both a market and a test. Thousands of Indian exporters, manufacturers and service firms sell to buyers abroad who are starting to ask AI assistants for suppliers. Many of those firms have thin or inconsistent information online, so they are easy for a model to overlook. India also has many languages, which makes it a useful place to learn how AI answers change from one language to another. If we can solve it here, the lessons travel.
Why build software rather than simply run a consultancy?
Palhade: Consulting is how we learn what works. Software is how we make it repeatable. Most businesses run visibility, prospecting and sales conversations in separate tools that do not share anything. We are trying to connect them, so a company can see how it is represented, find the right buyers and talk to them from one picture of the customer.
Many firms in this space guarantee placements in AI answers. You do not. Why?
Palhade: Because the answer belongs to the model, not to us. In this field a guarantee is either a guess or a misrepresentation. I would rather lose a deal than win it with a claim I cannot defend.
What should a chief marketing officer do this quarter?
Palhade: Ask the assistants your buyers use about your category, in every language and market you sell in. Write down who gets named, how you are described, and which sources are cited. Then compare it with how you describe yourself. The gap between those two pictures is your new marketing problem. For most companies it is bigger than they expect.
Are you still writing fiction?
Palhade: Not right now. But I still think like a novelist. Every product starts as a story about a person with a problem. The software is the ending we are trying to earn.
