In this episode of the Sunny Ray show, host Sunny Ray talks with Parag Vaish of NextNow about the gap that opens once companies move past off-the-shelf AI tools like ChatGPT, Claude and Copilot. Parag argues these tools will become table stakes, like Excel or Salesforce, so lasting differentiation has to come from custom capabilities built on reusable components. He traces lessons from Microsoft, StubHub, Tesla and Google, including Elon Musk's push for one-tap car buying and the trust trade-offs of building a weapons detector where misses can cost lives. He explains how NextNow treats each product as a set of ingredients, so its 51st build was more than 70 percent reusable, and shares a story about mining school board videos on YouTube to find districts in market. The conversation covers why mid-sized companies without in-house engineers and private equity owners focused on EBITDA are the best fit, how the team plans to stay small while scaling, and what separates companies that get AI right over the next five years. Parag also shares where to find NextNow's project case studies.
Off-the-shelf AI is becoming table stakes, so Parag Vaish explains how custom, component-built AI creates the differentiation big consultancies overlook.
What is the one thing a listener should walk away with from this episode?
Many companies stop at off-the-shelf tools like Claude, ChatGPT and Copilot, and those are wonderful. But there are remarkable things you can do beyond them. The real question is how you get differentiation in the long run, when those tools become as ubiquitous as Excel, Google Sheets and Salesforce.
What did building Tesla's in-car buying experience teach you about removing friction?
Elon wanted a one-tap car purchase. We built it, and he said we may have gone too far, so we added a couple of taps back. The bigger lesson was non-digital friction. Making the best car standard, with a sunroof and premium sound in every car, removed store visits and simplified supply chains and the factory too.
What did building an AI detector that must never miss teach you about trust?
At StubHub it wasn't life and death, but here we were detecting guns, bombs and knives. Accuracy was very high, but false positives like a laptop hinge are a real cost. We let the buyer choose a sensitivity level, so the decision and responsibility sit with them. You build the best product and create an environment for that choice.
When did you realize off-the-shelf AI tools would never be enough for most mid-sized companies?
Excel, Google Sheets and Salesforce are table stakes, and nobody says they beat a competitor because they are better at Excel. ChatGPT and Claude will fall into the same camp. Once I saw that, the question became how to be better than competitors. At NextNow we seek 10x opportunities, either ten times more output or ten times less cost.
How can your 51st build be 70% done on day one?
We look at every product through the lens of components, like ingredients in a kitchen that you recombine into different recipes. Front ends are easy now with tools like Lovable and Bolt. The ingredients form the foundation, and we are finding more than 70 percent is reusable from the first 50 builds. That lets users get 100 percent of what they want.
How did your team come to look at YouTube for 8,000 school districts?
We use first principles on every project and ask how we would do it differently in the presence of AI. To find who is in market, the school board meeting is where districts discuss needs before an RFP. Minutes were not helpful, but then we found video links, and about 8,000 of 13,000 districts have their videos online.
Which companies need NextNow most, and why are private equity owners a natural partner?
The best fit is companies without in-house engineers who know the latest models, often mid-sized firms stuck with the roughly 70 percent that off-the-shelf tools give them. Private equity owners care about EBITDA, and AI is the best lever for it. Most mid-sized firms lack engineering talent, so motivation and our ability to execute line up.
What does a mid-sized company that got AI right look like in five years?
They continue to innovate on top of the latest tech. Five years will feel like five lifetimes of change, so they build the muscle of harnessing new technology and engage partners like us to help. They recognize they must stay on the cutting edge to stay ahead of competitors. The companies that don't will meet their demise.
Building something daring? Sunny talks to founders like this every day. Fifteen minutes to see if your story belongs on the stage.
Claim your pre-interviewSend a founder here and we will know it was you.