ICONIQ

Shyam Sankar, CTO of @PalantirTech thinks the scarcest and most valuable resource in AI is taste. The possibilities builders now have at their fingertips defy imagination, and those who know what to build with them will define the next decade.

@ssankar

Disclaimer: bit.ly/3H4dQj0

The September edition of ICONIQ Pulse is live, featuring @ezCater joining @Uber, @enveda Series E, @WeAreLegora $200M ARR, State of Scaling, Frontier sessions, and more.

Subscribe and share: linkedin.com/pulse/ezcater-…

Disclaimer: bit.ly/3H4dQj0

You can prompt a lot of things. Taste isn’t one of them.

On October 8, @ConductAI CTO Henry Thompson and Sof Rami, a technical member of @AnthropicAI, will sit down for the first Conduct After Hours in New York to talk about the judgment behind building software with AI.

We'll be moderating the conversation, followed by drinks with engineers, founders, and operators across the city.

Apply to attend: luma.com/conduct-jlhl

Disclaimer: bit.ly/3H4dQj0

Two ICONIQ companies, @ezCater and @Uber, are joining forces to help shape the next chapter of workplace food.

ezCater brings deep enterprise expertise. Uber brings global scale and delivery infrastructure. Together, we believe they can create a more complete way for businesses to feed their teams.

Congratulations to @StefaniaMallett, @NihadRahman82, @dkhos, and your teams. We’re grateful to have partnered with both companies and look forward to what comes next.

Read more: bit.ly/4y8mJgr

Disclaimer: bit.ly/3H4dQj0

The hardest problems in AI may be the ones humans barely think about.

Folding laundry. Pouring coffee. Turning a sock inside out.

At ICONIQ Frontier, @physical_int co-founder @svlevine joined ICONIQ’s @rbernshteyn to explain why these seemingly simple tasks remain extraordinarily difficult for machines, and why general-purpose models could change that.

Physical Intelligence is working toward a shared intelligence layer that can learn across robots, tasks, and environments rather than engineering each system from scratch.

A few moments from the conversation:
04:22 Why the “easy” things humans do are so hard for AI
06:47 How generality changes the robotics problem
10:38 How training data evolves as robots become more capable
15:01 The intelligence layer as a kind of operating system for robots
20:23 Why robotics may not have a single “ChatGPT moment”

Disclaimer: bit.ly/3H4dQj0

The fundamentals of building a great software company still matter. It's the pace and discipline of some of the strongest companies that's setting a new standard.

The Great Sorting is well underway. Where companies land could depend on how quickly they adapt.

Read the full report: bit.ly/4hcZT0P

Disclaimer: bit.ly/3H4dQj0

Great used to look different.

For 15 years, software benchmarks were built around predictable growth decay, increasing expansion revenue, and improving efficiency.

Those benchmarks still matter. But they don’t fully describe today’s identified AI-native “Pacesetters,” who are reaching $100M ARR in quarters, growing 3–5x faster than the broader market, and sustaining 70%+ of gross new ARR from new logos deeper into their growth journey.

The ICONIQ Pacesetter Index benchmarks this cohort on the metrics that increasingly matter in an AI-native market, from growth and retention to gross margin and burn multiple.

What does great look like in 2026? The Index offers a clearer view.

Read the full Pacesetter Index: bit.ly/4d3vqkA

Disclaimer: bit.ly/3H4dQj0

Pacesetters spend up to 3x more than peers at the same stage.

Yet their burn multiples remain at parity with peers, or better, as that higher investment generally translates into more revenue generated per dollar spent.

See the full data: bit.ly/4hcZT0P

Disclaimer: bit.ly/3H4dQj0

We published the ICONIQ Pacesetter Index a few weeks ago, and the reactions since have reinforced just how quickly software benchmarks are moving.

Some of the discussion focused on the widening gap between the companies pulling away and the rest of the market, and put it in historical context. A $100M+ ARR company growing 115% used to be a rare outlier. In the Pacesetter cohort, that's the median.

Other reactions dug into the economics behind that growth, including strong gross margins and net revenue retention.

For founders and operators, the useful question is where your own company sits against these benchmarks.

Read the full Pacesetter Index: bit.ly/4d3vqkA

Disclaimer: bit.ly/3H4dQj0

How long can a conviction about AI stay true when the technology keeps changing underneath it?

“Every three, four months you get a curveball,” @nikesharora told the ICONIQ Community at Frontier.

In conversation with Divesh Makan, the @PaloAltoNtwks chairman and CEO shared how his thinking has evolved as agents, open-weight models, and new approaches to inference have emerged, and why that pace of change should push founders to ask a harder question than where to add AI.

For some companies, the opportunity may be to rebuild the product around what AI now makes possible.

Nikesh also discussed why startups can still have an edge in new categories and how AI is changing cybersecurity by compressing the time it takes to find and exploit vulnerabilities.

A thoughtful opening to ICONIQ Frontier on what remains unsettled, and what that means for the people building through it.

Key moments:
03:21 Why AI assumptions can change every few months
10:07 Rebuilding products from first principles
16:42 Why startups can win new categories
19:28 How AI is changing cybersecurity
27:21 What Nikesh looks for in founders and new ideas

Disclaimer: bit.ly/3H4dQj0

On October 2 @outtake_ai and @OpenAI are teaming up to share how they created Dynamic Discovery to help power a more adaptive approach to finding threats.

Outtake CEO @adylon7, Outtake engineer Tracy Meng, and OpenAI engineer Nishanth Singaraju will unpack how they built this technology using OpenAI’s Agents SDK and how they believe adaptive search represents the next evolution of digital risk protection.

Register here: outtake.ai/openai-x-outta…

Disclaimer: bit.ly/3H4dQj0

We’re honored to partner with @viswacolluru and the @enveda team as they accelerate the discovery of medicines for people in need around the world.

Viswa grew up seeing firsthand how medicine can change lives, and set out to find more of it, faster.

Some of the world’s most important medicines, including morphine, aspirin, and metformin, originated in nature. Yet an estimated 99% of nature’s chemical diversity has never been examined. Viswa founded Enveda in 2019 to explore that untapped potential.

Its AI platform, PRISM, identifies promising molecules already at work in nature and helps determine what they could treat. Seven years and 17 development candidates later, three of Enveda’s medicines are now in human trials, with two reporting positive early clinical results in 2026

businesswire.com/news/home/2026…

Disclaimer: bit.ly/3H4dQj0

Wealth management runs on a zoo of point solutions that do not talk to each other.

@MarkSwan_ built @neviswealth deeper instead, straight into the custodians. Press a button and the account is opened.

Learn more about Mark and his journey to co-found Nevis: youtu.be/hTLZA0Eq1gk

Disclaimer: bit.ly/3H4dQj0

Legora has crossed $200M ARR.

In our 2026 State of Scaling report, we highlighted the company’s climb from $1M to $100M ARR in just 18 months. The next $100M came in less than six.

Since we partnered with @MaxJunestrand, @siggelabor, and the team a year and a half ago in their Series B, @WeAreLegora has grown to 130,000 monthly users across 2,100 firms and legal teams in 80+ countries, with the US now its largest market by revenue.

It’s been a privilege to watch this team build with such speed and rigor. Congratulations to everyone at Legora on an extraordinary milestone.

Legora is hiring: legora.com/careers

Disclaimer: bit.ly/3H4dQj0

Anthropic, Legora, Sierra, ElevenLabs, OpenAI, and Glean are reaching $100M ARR in one to three years, a milestone that used to take software companies five or more.

The growth curve for AI-native companies looks fundamentally different.

More on what sets the outliers apart: bit.ly/4hcZT0P

Disclaimer: bit.ly/3H4dQj0

What separates the software companies pulling ahead from the ones falling behind? @alexrkonrad dug into our latest research and put it plainly. His piece highlights some of the key signals we identified in our report: faster paths to $100M ARR, sustained new-logo growth, aggressive investment, and improving burn at scale.

Download our annual Topline Growth and Operational Efficiency Report: bit.ly/4hcZT0P

@UpstartsMediaCo

Disclaimer: bit.ly/3H4dQj0

The Great Sorting has begun.

A new class of software companies identified from our dataset, which we call Pacesetters, is compressing years of growth into a fraction of the time. They’re spending more, growing faster, and generating more for it.

Earlier this year, we explored this shift in “A Coming Age of Reason.” Our latest report puts the data behind that thesis.

Download the 2026 State of Scaling: our annual Topline Growth and Operational Efficiency Report: bit.ly/4hcZT0P

Disclaimer: bit.ly/3H4dQj0

A 20-kilowatt satellite is flying today, and a 100-kilowatt platform is planned next.

@neelkunjur, co-founder of K2 Space, on why progress in space compounds when improvements in launch enable improvements in spacecraft, and why K2 builds for mass production rather than one-off engineering.

Learn more about @KaranKunjur and Neel's journey building K2 Space: youtu.be/9VVMy7QmHBY

Disclaimer: bit.ly/3H4dQj0

Opening custodial accounts can take a wealth management firm months. @neviswealth does it the same day.

@MarkSwan_ started with one of the industry's biggest bottlenecks, the manual PDFs behind every new account.

Learn more about Mark and his journey to co-found Nevis: lnkd.in/gCGgP-q3

Disclaimer: bit.ly/3H4dQj0

"Every engineer you talk to wants to prove those people wrong."

@neelkunjur, co-founder of @K2SpaceCo, on why some of the best engineers seek out problems others call impossible, and how a founding group of engineers created a cycle of talent that keeps compounding.

Learn more about @KaranKunjur and Neel's journey building K2 Space: youtu.be/9VVMy7QmHBY

Disclaimer: bit.ly/3H4dQj0