Maverick Intelligence

Has a friend ever doubted your business idea? I recently interviewed Maverick F., co-founder of AI startup Maverick Intelligence. A friend told him he couldn't build an AI startup because he didn't have a technical background and therefore wasn't cut out to be a founder (ouch), but he chose not to believe him. His own journey is proof that the only ceiling is the one you accept. Exactly one year after that conversation, he secured his first investment from Boost VC for his startup, Maverick Intelligence. The platform identifies who is on your website and turns anonymous visitors into leads with a single click. Maverick is a Cambridge graduate, former Model ML operator, and one of the most relentlessly self-driven young founders in UK tech right now! His perspective is commercially sharp, brutally honest, and deeply energised by the people who said it couldn't be done. In this episode, we cover: 1. How to build an AI company without a technical background 2. How to turn doubt into fuel 3. Why cold outreach, not AI agents, is still what actually closes sales deals The barrier to building has never been lower. Has a friend ever doubted your business idea? Watch the podcast on YouTube: https://lnkd.in/exxgFxpr What the podcast on Spotify: https://lnkd.in/ez2-erTv

Knowing whose agent visited your site is infinitely more impactful than knowing which human visits your site. That sounds wild, but it’s incredibly logical. Here’s why: 1. Google Ads has been gamified. LLMs have not. In competitive markets, the most common AdWords strategy is to bid on the keywords of your competitor. For example, if you google Harvey, the top 5 results will all be competing legal AI companies. By contrast, for model providers, It is against their interests to provide gamified information. They must provide the most relevant information to the human, at all costs. Aravind Srinivas, founder of Perplexity, spoke about this when justifying his decision to remove advertising on the platform. Relevance is the product. 2. Google has no context. LLMs do: Traditional search engines have no context on your search. By contrast, LLMs have significant person-level context. The agents from LLMs will automatically search the most relevant sites. This increased level of context means the agent’s searches are infinitely more thorough and personalised. They search the sites which are most relevant not just to a 5 word search, but to your entire individualised context. 3. We trust two brains infinitely more than we trust one: With Google search, it's you and google versus the list of searches. You do all the decisioning. With LLMs, it's you, your context, and the LLM's interpretation of your search. As a result, our level of trust in LLMs to influence decisions is way greater than our trust in any other search modality. Our level of trust in the models is only going to increase. Exponentially. With Maverick Intelligence, we're building the world's greatest knowledge graph on human-to-agent data. The value we saw from our first few months deanonymising human site visitors was immense. The value we're seeing from deanonymising agent visitors is groundbreaking.

• 8 billion humans. • 80 billion agents incoming. The new Maverick Intelligence website is now live. Check it out in the comments below. 🌎

We've been working with Giorgio and the Thrad team for less than two months now, and their wins have been epic. • 1400 website visitors identified. • 120 qualified prospects identified. • 16 meetings booked. Thrad is a San Francisco–based ad infrastructure company for AI conversations. Publishers use Thrad to monetise with in-chat ads, and brands buy ads in AI. Their clients include the likes of MongoDB, Turk Telekom, and Binance. Giorgio Toledo, it's a privilege to work with you and the Thrad team.

Turning human website visitors into customers is so 2025. We spent the past 4 months building the most accurate, deterministic web visitor identification technology on the planet, for humans. But we always knew this would be our first step on a path towards an infinitely greater vision. An agentic vision. Today, AI agents make up 53% of all online searches. By this time next year, I'm confident that percentage will reach 99%. Yet currently, agent traffic gets classified as bot traffic. It isn't. It's orders of magnitude more valuable. More valuable, dare I say it, than human traffic. By building our own data pipelines from scratch with agentic search in mind, we've understood not just how humans reason through sites, but how AI agents reason through sites. We're answering questions like: • Which LLMs they come from? • What prompts bring agents to specific sites? • Where agents get stuck on sites traditionally built for humans? • Who are the masters of these AI agents? If you want to turn AI agents into customers, comment "agents" for beta access.

Last week, Nicholas Saremi and I launched an API and a Webhook for Maverick Intelligence. We only told our power users, as we wanted to see what they'd do with it. We were pretty astonished by what they built. Here are just some examples: • Personalised outreach via cross-referencing each web visitor with their stage in the CRM. • Ranking website visitors based on how likely they are to be ideal customers. • Using ICP website visitors as an anchor to get company-wide buy in. The Maverick Intelligence API and Webhook are now live. DM me for access, and comment below if you can guess what we're building next....