AI, Unscripted · No. 06 · Rishabh Jain, CEO, Fermat · Interview by Sean Simon at ShopTalk Fall ’26 · In partnership with RapidCanvas
Intent-based personalization is a different idea from the personalization most brands know. Instead of reordering products based on who someone is, it reshapes the entire website experience based on what the shopper is trying to do.
Rishabh Jain is the CEO of Fermat — named for the historical mathematician, a nod to his PhD in solid-state physics. Fermat uses AI agents to watch the full consumer journey, from the ad someone clicked to how they behave on site to their order history, then serves a site layout built to help them decide faster.
At ShopTalk Fall ’26, Rishabh and Sean Simon covered how intent-based personalization works, why he exited his entire sales team, and why he thinks this is the best moment in history to be an individual contributor.
Table of Contents
5 Lessons on Intent-Based Personalization and Running an AI-Era Company
1. Every Entry Point Signals a Different Mission
Channels have always converted differently. Rishabh’s framing is simple: someone arriving from Meta is usually on a discovery mission, while someone arriving from Google is on a decision mission. The website should cater to that mission instead of treating everyone the same.
Fermat’s agents are trained on how people shop. If a visitor spends a long time in the size guide, for example, they’re probably not finding the information quickly enough. Intent-based personalization uses those signals, plus order data, to change the whole site layout — not just the product order.
One important detail: it isn’t generating pages in real time yet. Fermat pre-generates modules and picks the most relevant ones when a visitor arrives. Truly real-time generation is still too slow.
2. Measure Behavior, Not Just Outcomes
Attribution, incrementality, and MMM all measure outcomes: does a dollar here drive incremental revenue? Rishabh argues there’s a richer question. Does a dollar here drive a certain type of customer or a new behavior?
If Meta spend drives discovery across a wider range of products while Google keeps selling your bestsellers, Meta may be more valuable than its dollar-for-dollar numbers suggest. That trained behavior shows up later, outside the window traditional measurement can see.
3. AI Helps Big Companies More Than Small Ones
Fermat started about four and a half years ago, after Apple’s ATT changes made identity data harder to use. The original idea was to use behavioral signals instead of identity. The first version was manual — essentially a landing page builder on steroids — and served smaller brands.
When AI arrived, behavioral data suddenly became usable at scale, and Fermat moved from SMB to enterprise. Rishabh’s take: AI’s benefit is much higher for large companies, because they have more data and can now move as fast as smaller ones.
4. Founder-Led Sales Can Scale Further Than You Think
The most controversial moment in the interview: Rishabh exited his entire sales team about three months ago and now handles sales himself — and the business is growing faster than ever. At large enterprise contract sizes, he says, buyers want to hear from founders and CEOs, and AI has collapsed so much of the prep work that he can manage around 25 active deals at once.
Sales cycles that used to run six to nine months — when deals closed at all — now take about three. No cold outreach; advisors and partners fill the funnel. To scale, he’s weighing two models: a McKinsey-style partner model with senior hires who carry trust with peers, and the approach Bret Taylor and Clay Bavor used at Sierra, where an orchestrated team builds around the founder’s vision.
People don’t buy from sellers at that contract size. They buy from CEOs and founders.
Rishabh Jain, Fermat
5. This Is the Era of the Super IC
Inside Fermat, AI has shifted how people are measured — from inputs to outcomes. Because anyone can now retrieve data and take action across org boundaries, an individual contributor can own an entire outcome. Rishabh says it’s become hard to convince people to become managers.
He even uses Meta’s new Muse agent to run his pipeline, pulling from his calendar, call recorder, email, and Notion. It sends a Sunday summary and preps him before meetings, and he says he can’t remember the last time he opened the CRM. Beyond work, he’s excited about AI helping people manage their health with longitudinal data, and about a single person being able to launch a real side business with an agent’s help.
Related on The MarTech Matrix:
About Rishabh Jain and Fermat
Rishabh Jain is the CEO of Fermat and holds a PhD in solid-state physics. Fermat provides intent-based personalization for brands and retailers: AI agents analyze the full consumer journey — ad source, on-site behavior, and order history — and assemble a website experience designed to help each visitor make a decision faster.
The company launched about four and a half years ago and has evolved from a manual landing page builder for smaller brands into an AI-powered enterprise platform. A customer recently described Fermat as a tailless e-commerce platform — handling everything on the front end except payments and order management.
Frequently Asked Questions
What is intent-based personalization?
Intent-based personalization tailors a shopper’s website experience to what they are trying to accomplish, not just who they are. Fermat uses AI agents to analyze the ad a visitor came from, on-site behavior, and order history, then serves a site layout designed to help that visitor decide faster.
How is intent-based personalization different from traditional personalization?
Traditional personalization often reorders products or swaps content based on identity data. According to Rishabh Jain, intent-based personalization changes the entire website layout and presentation based on behavioral signals, such as whether a visitor is on a discovery mission from Meta or a decision mission from Google.
Does Fermat generate websites in real time?
Not yet. Rishabh Jain explains that generating a full website on the fly is still too slow. Fermat pre-generates modules and selects the most relevant ones for each visitor as soon as they arrive, so the experience is rendered in real time but not generated in real time.
Why did Fermat’s CEO take over sales himself?
Rishabh Jain says enterprise buyers at large contract sizes prefer to buy from founders and CEOs, and AI has reduced sales prep to minutes. Since exiting the sales team, Fermat’s sales cycles have shortened to about three months and the business is growing faster than before.
Leave a Reply