Informed i’s Weekly Business Insights

Way back in the summer of 2017, AI researchers at Google put out a paper called “Attention Is All You Need,” in which they described a new type of neural network called a transformer. It proved to be very good at processing long sequences of data, especially text. 

Nine years on, transformers are the engines inside every major large language model on the market. “The entire AI industry is built on transformers.  But transformers are starting to show their age. Many of the recent advances in LLMs, such as the development of so-called reasoning models and their ability to handle large amounts of input at once, are not neat extensions of that core technology but workarounds that patch over some of its fundamental flaws.

A growing number of scientists and engineers are now asking what’s coming next. LLMs are not going anywhere, but the way they get built is up for grabs.   Enter a wave of startups hoping to push the boundaries of this boomtown technology. Some will no doubt fail—but they have everything to play for and far less to lose than the companies at the front of the pack today.

But first, the problem. The key strength of transformers lies in a mechanism called dense attention, which encodes the meaning of a block of text in a series of numbers. The process involves comparing every word (or part of a word, known as a token) in that text with every other word via a form of multiplication.

Dense attention can capture the meaning of text with remarkable accuracy. But as the length of that text grows, the number of computations needed to process it adds up fast. A document 10,000 words long might require a transformer to perform 50 million multiplications. That’s the main reason LLMs suck up so much power.

What’s more, transformers struggle with what many of the latest models are designed to do. Because of the way they process text word by word, transformers are not great at keeping track of a lot of information at once (in other words, what’s known as their context window cannot get too large). And yet if LLMs are to carry out harder tasks, they will need to take in larger amounts of data: a whole library of documents, an entire code base, or in the case of agents, output from other LLMs.  As for reasoning models, they work by writing notes to themselves (in a kind of scratch pad known as a chain of thought) and then reading them back, which again adds to the amount of data to stay on top of.

As LLMs get bigger and better, transformers have become a bottleneck. The technology’s key strength is now a limitation.  Here are four new ideas for how to solve the transformer problem—innovations that could change LLMs for good, making them faster, far more efficient, and (maybe) even smarter.

Rethinking attention.  An obvious way to make LLMs faster and cheaper is to tackle the problem head on and change the way attention works. Swapping out dense attention for a mechanism called sparse attention, which runs calculations on only some pairings of words in a block of text instead of all of them, can radically reduce the amount of computation LLMs need to do. Subquadratic, a startup based in Miami, claims it has invented the first sparse attention mechanism that rivals top mainstream LLMs on a handful of tasks, including search and coding. It’s a huge claim (and some people in the industry remain skeptical).  It has developed a mechanism it calls power retention, which stores only the most relevant information for a given task and ensures that the amount of data an LLM has to keep track of doesn’t blow up.

Making models smaller and more flexible.  Liquid AI, an MIT spinout based in Cambridge, Massachusetts, hasn’t changed or ditched transformers fully but pairs them with its own tech, liquid neural networks, to build what cofounder and CEO Ramin Hasani calls LFMs (liquid foundation models).  Liquid AI’s models are far smaller and use less energy than most LLMs. The firm builds models for car makers, including Mercedes, which run on the small chips inside vehicles.  Liquid neural networks were inspired by worm brains. They are an extension of another type of neural network that predates transformers, called convolutional networks. The key innovation is a mechanism that lets a model adapt its behavior to new information, so it can learn as it goes. That’s not possible with transformers: Once a model is trained, its behavior is fixed.  Liquid AI’s first models were pretty basic but could fly drones or drive vehicles.  Hasani thinks transformers were just the beginning: “Your brain is an AGI system, you know, and it operates with 20 watts of power. How is it possible? We can get a lot more innovative.”

Generating text all at once.  Almost all LLMs produce their output one word at a time. It makes sense, because that is how people speak and write. But for computers, it’s very inefficient.  It is faster and cheaper for LLMs to generate text all at once—spitting out whole sentences or paragraphs in one shot. That’s the approach taken by Inception, a startup based in Palo Alto, California, which is building LLMs using a technique called diffusion.  Diffusion is better known as the technology that drives most image and video generation models. Diffusion models are trained to take a random grid of pixels—like the static on an old TV set—and turn it into an image. They do this by working on all the pixels at the same time, figuring out which need changing to make the static look more like a high-definition photo.  It turns out this process works on text too.  Inception is not the only company betting on diffusion.

Moving beyond words.  Pathway, another startup based in Palo Alto, is perhaps the most extreme of this new bunch. It wants to free LLMs from the constraints of language.  The firm has built a type of LLM called Dragon Hatchling (named after the dragons in Terry Pratchett’s novel Color of Magic, which materialize if you think about them hard enough). Its standout result so far is a high score on a benchmark that pits LLMs against more than 250,000 very hard sudoku puzzles. Dragon Hatchling beat more than 97% of the puzzles; several leading LLMs from the top labs failed to solve any.

AI is shattering many decades-old assumptions about how growth happens. The fundamentals of growth still matter, but the way companies create, capture, and scale growth with AI has fundamentally changed the game. 

To narrow the gap and turn AI into a force multiplier for growth, CEOs need to fundamentally rewire how they operate to move faster, more iteratively, and with clearer direction. That starts with rejecting the myths that can sink their growth aspirations.  According to the authors, in their experience, they see seven common myths that can hamper business leaders’ ability to turn AI into value.

Myths that limit ambition

1. Myth: Above-market growth goals are enough.  Reality: AI raises the ceiling on how much growth is possible. While nearly 75 percent of companies set above-market growth targets, that’s not bold enough.  Research finds that leaders who set bold growth ambitions—underpinned by AI—can unlock nearly three times more value than they imagine.

2. Myth: AI is primarily a productivity tool.  Reality: Some 80 percent of companies use AI to improve efficiency, but AI can do much more. The companies creating the greatest value are using AI to improve both efficiency and effectiveness.  Greater effectiveness creates new demand, increases conversion, and accelerates growth and innovation.  

Myths that preserve old ways of working.  

3. Myth: Layering AI onto existing processes drives growth.  Reality: Bolting AI onto existing processes—often resulting in a proliferation of isolated use cases—doesn’t lead to growth. Rather, it is activity without architecture. AI high performers are three times as likely to say their organizations have fundamentally redesigned individual workflows.

4. Myth: The chief technology officer (CTO) should lead the AI transformation.  Reality: The approach associated with growth is to undertake a business transformation enabled by technology, with people at the center. That’s why leaders at half of top-performing companies co-create strategy across technology and business to redesign new pathways to growth.

Myths that slow down value capture

5. Myth: Scaling AI is mostly a tech rollout.  Reality: Creating enterprise-wide growth involves more than new technology; it involves new ways of working, capabilities and skills, and governance models. Achieving this degree of change requires resources dedicated to change management. The rule of thumb is that for every $1 a company spends on an AI initiative, it needs another $3—and sometimes more—for change management.

6. Myth: AI initiatives take a long time to create P&L value.  Reality: In our experience, AI leaders often see impact on leading indicators within weeks and create substantial economic value within three to six months. That happens when companies focus on high-value commercial workflows in priority domains and iterate rapidly.

7. Myth: Companies need perfect data in order to introduce AI.  Reality: Good data matters, but in our experience, 80 percent of companies do not need to build a data lake or rebuild their entire data architecture to get value. “Good enough” data is sufficient to start capturing value quickly.

Business leaders who want to shake off the growth myths and build a different kind of commercial growth engine can do this by taking concerted steps:  identify and test growth opportunities, build and iterate AI in execution, and scale AI and capute the value.

What CEOs can do now.  Developing an AI-driven growth engine is an investment of time and resources. But the authors’ experience has shown that companies can make progress quickly, especially if they tackle three actions first:

See the full set of AI-enabled growth opportunities in days, not months. Align on specific outcomes and commit to bold goals. And Mobilize cross-functional teams and commit the resources to fuel them.

Nissan was desperate. Merger talks with Honda had fallen apart. The company was in the worst crisis of its nine-decade history, and the board needed someone to lead it out.  Its choice: Ivan Espinosa, just 46, and not Japanese—unusual in a country where corporate leadership has long been considered a job for a Japanese national. Espinosa remembers being shocked by the request.

The CEO jokes that he was “born in Nissan.” He was, in fact, born in Mexico, where he started with the company as a product engineer in 2003. After roles in Southeast Asia, Europe, and Latin America, he moved to Japan in 2016 and became Nissan’s chief planning officer in 2024, exposing him to all the ways the company failed to right the ship. “I knew what had to be done,” Espinosa told Fortune earlier this year. “It was obvious you had to resize the company.”

Nissan’s turnaround is about more than just whether the 93-year-old carmaker has a future.

For decades, global carmakers like Toyota, Nissan, General Motors, and Volkswagen manufactured and sold their products all over the world. But that model no longer fits today’s more protectionist, more competitive world. Nissan’s current strategy underscores this transition, as the company orients itself around two markets: China and the U.S.

“You have a China ecosystem, and you have the U.S. ecosystem,” Espinosa said. “If you want to be a global company, you need to live in both.”

For years, Nissan had suffered from overproduction, high costs, and slow product development.  

Nissan’s then-CEO, Makoto Uchida, was out, and the Mexican-born Espinosa took his place.  Within six weeks, he’d developed a plan to both cut costs and keep up with the competition. The result was the Re:Nissan plan, announced in May 2025. The plan promises 500 billion yen ($3.1 billion) in savings; seven plant closures; 20,000 layoffs; and a reduction of development cycles to a little over two years.

Nissan now expects to return to profitability in the current fiscal year. But “it’s easy to cut costs,” Nakanishi says. “It’s more difficult to restore the value of the brand.”

Paradoxically, Nissan may be fortunate in that it’s going through its own painful restructuring ahead of its competitors. Honda posted the first annual loss in its history last fiscal year, and has gutted its EV plans. In April, Toyota abruptly replaced its CEO, Koji Sato, with CFO Kenta Kon after reporting a $9 billion hit to profits as a result of Trump’s tariffs.

Carmakers used to be the ideal global manufacturer, running vast cross-border supply chains and dominating large markets like China and India. Much of that swagger has faded as legacy carmakers are buffeted by the EV transition, new Chinese competition, and supply-chain pressures including Trump’s tariffs, rare earth export controls, and an AI-driven memory-chip shortage.

Nissan might be turning into something akin to a “regional multinational,” sharing technology and platforms across different markets, but moving away from a single global model. In North America, it must localize production and rebuild trust among dealers; in China, it needs to move at “China speed” and export domestic innovation to the rest of the world.

Today’s most successful brands are no longer competing on products alone; they’re creating interconnected ecosystems that keep customers engaged over time. By pairing products with relevant services, communities and experiences, companies are fostering deeper relationships that extend well beyond the point of purchase.

As this shift reshapes consumer expectations, retailers and brands alike must rethink how they deliver long-term value and loyalty. Here, Forbes Communications Council members discuss how ecosystem-driven strategies are changing retail, and what these changes mean for the future of the industry.

1. Building Relationships Beyond The Product.  Leading brands are moving beyond selling products to creating connected experiences that solve broader customer needs. Products can be copied, but ecosystems built on trust, value and ongoing engagement are much harder to replicate. For retailers, that means focusing less on individual transactions and more on building long-term relationships.
2. Prioritizing Consumer Retention Over Transactions.  Brands are shifting focus from product to consumer to meet changing market demands. With so much product redundancy, consumers have more options to choose from than ever before, opting for the organization that best understands their nuances and tailors to their needs. As a result, the sales role is adapting as consumer retention becomes the biggest priority in retail.
3. Creating Value Throughout The Customer Journey.  The strongest brands are no longer competing on products alone. They are building ecosystems that remove friction, anticipate customer needs and create value beyond the transaction. AI accelerates this shift by connecting data, experiences and services. The future of retail belongs to brands that become trusted partners throughout the customer journey, creating continuous value rather than simply enabling transactions.
4. Connecting Commerce, Service And Community.  Leading brands are moving from selling products to orchestrating trusted experiences across commerce, content, service and community. In retail, the differentiator becomes context: knowing who can act, what data is used and how experiences connect. The future won’t be won by more channels, but by ecosystems that feel seamless, accountable and genuinely useful.
5. Creating Seamless Cross-Channel Experiences.  The tactical priority is continuity: preserve identity, intent and attribution as customers move from paid or organic discovery to web, app and even in-store. Making every channel smarter because it knows what happened in the last one is the goal.

The other ways are:

Extending Value Beyond The Purchase

Turning Products Into Ecosystem Entry Points

Inviting Customers Into A Brand Experience

Designing Around The Consumer Journey

Replacing Transactions With Lasting Relationships

Remaining Useful After The Sale

Creating A Sense Of Belonging

Measuring Success Through The Customer Lens

Building Ecosystems Around Customer Habits

Creating Reasons For Customers To Stay

Removing Friction Through Strategic Partnerships

And Making Ecosystems A Competitive Advantage

The most valuable thing a professional produces is not a faster analysis or a better summary. It is insight: a genuinely new way of seeing a problem, a connection that changes how they understand a situation, or a pattern that nobody has named. 

What makes this kind of discovery hard is that expertise, the very thing that makes professionals effective, also makes certain kinds of insight difficult to reach. The frame that lets them see a problem clearly also shapes what they look for — and what they stop looking for. But insights that change thinking are often found at the edges of that frame, such as in the friction between competing interpretations.

Conversational AI is already moving in this direction. A well-constructed prompt can surface competing interpretations, expose gaps, and challenge assumptions. But agentic AI — systems that are configured and directed rather than conversed with — can take users further still. Unlike a prompted conversation that is bounded by what a human supplies and thinks to ask, an agentic system holds more data and sustains analytical orientations across entire data sets without losing the thread. The discovery moves are the same; the depth is not.

Using agentic AI this way demands a professional skill different from both prompting and automation: knowing how to design systems for insight and how to make sense of what they reveal. The authors call it directing intelligence.

Four Approaches to Discovery With Agentic AI.  Discovery rarely arrives through a single, well-aimed question. It tends to emerge from friction: from putting things in contact that are normally kept apart. Each of the four approaches that follow creates a specific kind of friction: between competing interpretations, between data and the conversations that surround it, between causes and the levels where they hide, and between categories and the reality they were meant to describe. The insight emerges from the friction itself. 

1. Use Multiple Lenses.  Apply competing frameworks simultaneously, and read the contradictions. The same data set through multiple agents, each with a different analytical directive?  What does the friction between well-reasoned analyses reveal that no single analysis would find on its own?  Each move is also available through prompting. What agentic architecture adds, through what agents access, do, and pay attention to, is depth.
2. Surface Silences.  Compare what the data contains with what the organization discusses.   Interview transcripts, operational data, and field records mapped against formal documents and stated priorities.  What does the organization know but never name, and what does that silence cost?
3. Bridge Levels.  Trace a problem from where it surfaces to where it originates.  Data connected across every organizational level simultaneously.  Where does the real intervention sit, and why aren’t we looking there?
4. Stress-Test Categories.  Test your classification system against operational reality. Formal categories mapped against the full behavioral record.  What are our categories hiding — and what falls outside them entirely?

How to Direct Intelligence Skillfully.  Gaining the most useful results from directing AI agents to surface new insights requires practice, like any new skill. The following are some guidelines to keep in mind.  Configure for discovery, not answers.  Treat the unexpected as signal, not error.  Evaluate proposals, not conclusions.  And track what you advance, and what you reject.

Having learned to cook in his parents’ kitchen, Andrés trained under several of Spain’s top chefs before immigrating to the United States and opening his first restaurant in Washington, DC. In the three decades since, he has not only built a culinary empire that stretches from Los Angeles to the Bahamas but also founded a high-profile nonprofit, World Central Kitchen, that provides fresh meals on the front lines of crises. His latest cookbook, Spain My Way, came out this spring.  The following are some of the key insights from his interview.

Why did you become a chef?  My parents were both nurses who loved to cook. My mother was the Monday-through-Friday cook to feed the children, and my father was the weekend cook to feed more people. My brothers and I would help in the kitchen, or maybe it was bothering that only looked like helping.  I hadn’t been doing well in school, and a private culinary school opened nearby. They took me because they needed students. But because it wasn’t yet finished—there was no kitchen—I first went to work in some of the best restaurants in Barcelona, and I never stopped. I’ll always find a way to be busy cooking.

What lessons did you learn from your famous mentors like Ferran Adrià?  When I met Ferran, he was 23 or 24, already manning El Bulli, and becoming known for pushing boundaries. Very late one night, he came to eat in the restaurant where I was working, and I made him gambas al ajillo, or garlic shrimp, and my fate was sealed. I began working in his kitchen, where over the next decades he reinvented the restaurant and created a place where everyone had to go. He taught me that you can’t be a good chef if you don’t love to eat. And you become a great chef by wanting to pass that love through your cooking to others. He was always reading books about traditions, dishes, great chefs, and restaurants of the past, but at the same time, was the guy who said, “When we master this, we need to forget about it. We should not be settling for what we are taught. Nothing is so sacred that we cannot change or improve it.”

Why did you decide to immigrate to the United States and make Washington, DC, your home base?  Life has a path for all of us. Some of us fight it. Some of us follow, and that was me. I had always wanted to come to America. But first, there was a boat: the Juan Sebastián de Elcano, a majestic tall ship of the Spanish Navy. I enrolled in military service and requested to go to that boat, but instead I was assigned to cook for an admiral. After a few months of getting to know him and his family, I finally opened up and said, “Hey, I love being here, but I want to go on a boat—on one boat in particular.” And he let me go, which brought me to Africa and the Caribbean and America.

How do you do that? What kind of leader are you?  I think I’m a leader in the making. The more I know, the more I know that I know nothing. There is no perfect recipe that works every time in every circumstance for every person. It’s also important to understand that our business is one of the most difficult to be in.

As you expanded your business, first in DC and then to other cities, how did you create distinct and innovative restaurants under one consistent José Andrés brand?  The early days were easier because I had all the restaurants near where I lived and worked, so I could walk from one to the other. That was very fulfilling because in two hours, I could spend 15 minutes in each and then decide which one to give more time to that day. And, while we opened near an arena and office and government buildings and the Shakespeare Theater, I watched the restaurants create a neighborhood. There are examples all across America of chefs going to a place where not much was happening and, over two or three years, creating new energy and building up the town so that everybody wins. For me, new restaurants are a way to learn other stories.  Now if you ask me, “Do you want to go back to the happy days of having only one or even five restaurants in a neighborhood?” I would probably say yes.

Which talents and skills did you bring from your culinary career to the humanitarian work of World Central Kitchen?  Jaleo is right across from the Clara Barton Missing Soldiers Office Museum, and Barton’s founding of the American Red Cross was always an inspiration. DC Central Kitchen—which not only collects food that would otherwise go to waste and serves it to the homeless but also hires people who are living on the streets or coming out of jail and trains them to be cooks—was also very important to me. Its founder, Robert Egger, once told me that philanthropy is not about the redemption of the giver but the liberation of the receiver, and I saw how he used food to create better lives.  When a big earthquake hit Haiti in 2010, I was in the Cayman Islands, and maybe because I was so close by and felt so powerless against the devastation, I said, “I’m going.” I went and began cooking in a few camps and realized that, while it’s not easy to feed people in a disaster, it’s also not so difficult. There is always food somewhere, and there are always volunteers, chefs, cooks, caterers, warehouses, and distributors. So I brought this mindset that the restaurant community is the world’s biggest and most powerful. We are all one phone call away from each other and with this network we can make sure that every problem is solved.  The other thing about restaurants is that nothing ever goes as planned. This is what the chef in me brought to the way we manage at World Central Kitchen: You have to adapt to whatever situation you’re in.

What qualities do you look for when hiring at your restaurants and World Central Kitchen?  I am always looking for people with a real passion for learning and for pushing beyond the horizon. You can teach skills, but you can’t teach passion.

Being in DC, you meet people from all around the country and the world with different backgrounds and opinions. You’re surrounded by diversity, so you understand the importance of building longer tables for everyone to sit at. If you don’t agree with someone, invite them to dinner and talk about it, and maybe you will learn to see eye to eye. As an owner of restaurants, a leader in nonprofits, a dad and husband, an immigrant and an American, this is what I believe: If we can all focus on building longer tables, not higher walls, we will make a better world for ourselves.

Last year, Ali Ansari was a master’s student at Stanford University when he sent a message to Andrew Maas, his computer science professor. Maas, who taught CS 224S: Spoken Language Processing, had earned his PhD from Stanford in 2015 and then worked as an engineer at Apple. Though Ansari was then only 24 years old, he had a proposal for his well-respected professor: join his fledgling startup, Micro1, as VP of engineering.

Maas didn’t respond at first, but that didn’t discourage Ansari. “I applied to the Stanford master’s program while I was building Micro1 pretty much for the sole purpose of getting into these research circles,” Ansari says. He was persistent, following up with Maas a few times. One year and several conversations later, Maas finally joined Micro1 as employee number 80.

This hustle encapsulates how Ansari, who is now on leave from Stanford, has built Micro1 into one of the fastest-growing companies in the AI training industry—one of the most critical and contested layers of the AI economy—landing the company at No. 37 on the Inc. 5000. Though it has pivoted a few times over the past few years, today Micro1 is a data lab and research partner that helps frontier AI labs improve their LLM models and agents. And now, at age 25, Ansari is the youngest CEO in Inc top 50 in an industry that can make 20-somethings billionaires in just a few short years.

At the center of Micro1’s growth is Zara, the AI recruiter that Ansari built while an undergrad at the University of California, Berkeley in 2022. Zara interviews and vets subject matter experts who apply to work as Micro1 contractors. It asks them highly specific questions to assess their level of expertise, ultimately aggregating the learnings of thousands of people with a particular expertise who work across a range of disciplines. Some of the job openings at Micro1 have included Sherpa language experts, international law experts, an audio engineer, and a scriptwriter.

Ansari says that one reason for the company’s growth and success so far has been its focus on the “human aspect” of data—recruiting, screening and onboarding experts to train the AI model, and paying them well, typically between $50 to $200 per hour, with some experts being paid well above that range.

But in the competitive and highly lucrative data-labeling space, the advantage that may best distinguish Micro1 is its ability to constantly reinvent itself as new opportunities emerge and a group of ambitious and well-funded rivals chase after them. That Micro1 has been able to survive, let alone grow at an impressive 6,963 percent rate over the past three years, suggests that the company’s expertise is not limited to the insights and information its many contractors pour into the model each day. It also suggests that Ansari, in the hunt to be one of the world’s youngest self-made billionaires, is a young CEO to watch.

Ansari founded Micro1 as an AI recruiter agent while an undergrad at Berkeley. At first, the company made apps and websites for clients. He started recruiting engineers, but felt the interviews took up too much of his time. So he created an AI screener to evaluate an engineer’s skill level. If the engineer passed the screener, Ansari would interview them at the final stage.

Starting in 2024, Micro1 slowly began shifting toward data after Ansari noticed the demand for engineers by one of its clients, a data-labeling company. He realized that this was something big when this client approached Micro1 to hire 600 engineers in three weeks. Ansari decided that instead of using his recruitment agent to vet engineers for other companies, Micro1 could simply become the company directly offering these experts to AI labs and capture the growing market. In early 2025, while at Stanford, he pivoted Micro1 from an AI recruiter agent to an AI training data infrastructure company, bagging the company’s first AI lab client.  And in early 2025, from his perch at Stanford, Ansari changed the trajectory of Micro1, beginning its meteoric growth, with annual recurring revenue jumping from $7 million in January 2025 to over $400 million this year.  Since then, there’s been no looking back.

Today, a few thousand candidates go through Zara each day, Ansari says, and about 20 percent are accepted. Zara has had more than five million signups and around three million interviews since its inception. “That’s only possible because of this autonomous agent that we built, which speeds up a lot of our data pipeline requests,” says Ansari.  Micro1’s advantage, explains Ansari, is its “best-in-class recruitment software,” human-first approach, and development of continuous evaluation as a service.

Despite Micro1’s stupendous rise, Ansari believes the biggest growth of the business is yet to come. He explains that there are two parts to its strategy. First is training: teaching AI models and robots new capabilities. This is where the company’s clients are mostly frontier labs and robotics companies—a vertical that’s slowly growing. The second is a product called Cortex, which focuses on AI agent evaluations.

Everywhere Jake and Michael, the founders of Nucleus Network, looked, talented founders were stuck not because their product wasn’t good enough, but because they couldn’t get in front of the right person. Meanwhile, Jake and Michael, were constantly asked, “Do you know anyone who…?” And they almost always did. Their ability to connect people wasn’t a side skill; it was the most valuable thing they offered. That realization became the foundation of a company built on one belief: the warm introduction is the most powerful force in business.

Their partnership works because they share a core philosophy: everything is figure‑out‑able. It’s the mindset that drives their company, their leadership style, and their approach to solving problems for founders. They don’t believe in waiting for perfect conditions; they believe in building momentum through action, judgment, and trust.

Use Warm Introductions as a Strategic Growth Engine .  The Nucleus Network operates across three interconnected pillars: Advisory, Ventures, and Community, all designed to connect the right people at the right moment and turn trust into opportunity. Advisory is their retainer “warm intros as a service” practice where venture-backed founders and agencies receive targeted warm introductions to customers, partners, talent, advisors, and more without ever sending a cold email. Instead of “spray and pray” outreach, Nucleus delivers curated, double-opt-in introductions that condense months of effort into a single trusted connection. Their clients rely on them not just for access, but for judgment, the ability to know who should meet, when, and why. 

Community is their invite‑only events arm, and it has become one of the most distinctive parts of the Nucleus ecosystem. They’ve hosted 100+ events across New York City, LA, Miami, San Francisco, Vegas, Chicago, and Europe, partnering with companies like Morgan Stanley, Cash App, MongoDB, Alliance Bernstein, Rho, Boardy, Sydecar, DraftKings, and more. These aren’t typical networking mixers; they’re curated experiences designed to deepen trust and create real relationships. Sauna and cold‑plunge sessions, tennis tournaments, backgammon nights, founder dinners, investor salons, and ecosystem meetups all serve the same purpose: get the right people in the right room and let trust do what it does. Their events have become known for their warmth, intentionality, and ability to create connections that actually lead somewhere. 

Together, these three pillars form a compounding system. Advisory drives introductions, Ventures opens doors to private deals, and Community deepens trust in person. Each pillar strengthens the others, creating a network that grows not through volume, but through quality.