Summary

Raanan sees cybersecurity in a “perfect storm”: geopolitical conflict has turned offensive cyber into a strategic weapon, capabilities keep drifting from states to criminals, and AI can execute attacks at unprecedented speed and scale. The escalation runs from analysis to human augmentation, workflow automation, and ultimately autonomous control. An AI agent could generate and launch thousands of infrastructure attacks simultaneously: “This is not a future thing. That’s today.”
AI will not merely improve cybersecurity; “it’s going to redefine it,” forcing enduring vendors to become AI-first and AI-native. The same model can identify a bank’s vulnerabilities for defenders or design an attack against it, while open-source systems such as DeepSeek remove the protection of a walled garden. For Wiz, Cyera, Island, and other platforms, AI therefore requires re-architecting the product—not adding another feature.
Cyberstarts’ results support a radically simplified seed thesis: select exceptional athletes before ideas, then help them find an urgent customer problem. Its $50 million first fund led investments in nine idea-stage teams; within three years, Raanan says the portfolio exceeded $25 billion in value and the $50 million had become close to $2 billion, producing Wiz, Island, and Fireblocks. Today Cyberstarts manages more than $700 million across five funds, with each of four active funds reportedly delivering over 100% IRR.
Raanan values demonstrated adversity above raw IQ because “startup is an insanely painful journey.” Fireblocks illustrates the wager: Cyberstarts invested $3 million in June 2018, then six of the company’s first 10 design partners went bankrupt during the crypto winter; founder Michael Chodorow’s history of navigating immigration, growing up without a father, and uncertainty helped him rebuild the plan. The key interview question is not what someone accomplished but why they made each consequential choice.
The Sunrise method reverses conventional product development by investigating pain before permitting founders to write code. Teams question dozens of CISOs, ask “Who’s the vendor you hate the most?”, and test claims against actual spending or workaround behavior. They tell customers a new team will spend about $100 million over three years on engineering and ask which problem they would want $10 million of that effort to solve off-balance-sheet, then may wait six months before building: “This is your last chance to pick the right problem.”
Wiz became the clearest proof that urgency, buyer simplicity, and enterprise pricing can overwhelm seemingly high entry valuations. After abandoning a secure-access idea that was important but not urgent, Wiz found a cloud-security product for which one CISO held the pain, budget, authority, and credentials needed to deploy it. It rejected PLG for large enterprise contracts and progressed through roughly $1 million, $2 million, $8 million, and $25 million of ARR in its first four selling quarters.
Cybersecurity is not a fixed market to map but “always a derivative of something else,” so Raanan concentrates on talent and process rather than predicting categories. He believes a new cyber company can surpass Wiz’s conditional zero-to-$32 billion, five-year outcome within a decade, given the market, evolving pain, talent, and some luck. His deeper counsel is to treat investing as a “life project,” grounded in real care for founders rather than profit generation alone.

Deep dive

1. AI turns cybersecurity into machine-speed conflict

In 2018, cybersecurity still looked like “a boring portion of IT”; today, Raanan sees global conflict making offensive cyber a lethal or strategic weapon. State capabilities continually drift toward criminal organizations and eventually “script kiddies,” creating a materially more dangerous baseline.
His warfare analogy progresses from better analysis to human augmentation, workflow automation, and finally AI control. Augmentation “might be enough” temporarily, but defenders ultimately need LLMs predicting attacks and AI agents able to respond in a timely manner. The Terminator future, he argues, “is reality today.”
Offensive tactics remain ROI-driven: inflict the greatest contextual damage with the least time, effort, and cost. Raanan’s specimen is attacking name servers supporting power stations, potentially depriving a country of electricity for days; an agent could produce and execute thousands of such scenarios simultaneously.
Dual use is unavoidable. A model trained on bank infrastructure can recommend defensive fixes or reveal how to take the bank offline, just as a drug model might propose therapies or poisons. Guardrails can be bypassed, and open-source models such as DeepSeek mean “AI is not walled garden anymore.”

2. Cyberstarts stripped seed investing back to the athlete

Leaving Sequoia in 2018, Raanan concluded that venture’s early-stage model was “completely broken.” Investors interrogated founders about markets, products, competition, and pricing when their answers would change within weeks—and the intended customer was absent from the conversation.
Cyberstarts instead committed to evaluating the person and “hunting for adversities.” From a shipping-container office in his backyard, Raanan started with a $50 million fund, led investments in nine teams without ideas, and made what many observers considered unusually large bets on people alone.
Three years later, he says those nine teams were worth more than $25 billion and the original $50 million had become close to $2 billion. The portfolio included cloud-security leader Wiz, enterprise-browser creator Island, $8 billion crypto-custody provider Fireblocks, and several companies acquired for hundreds of millions apiece.
Raanan attributes the hit rate to a narrow thesis, founder selection, Sunrise’s pursuit of product-market fit, committed partners, and luck. Cyberstarts now manages more than $700 million across four seed funds and a continuation fund; its roughly $45 billion portfolio represents, by his calculation, 50% of private cybersecurity’s worldwide market capitalization.

3. Adversity is the founder signal Raanan trusts

High IQ helps, but Raanan’s strongest founder signal is “the adversity and the ability to overcome it.” Startups reliably produce pain, uncertainty, and broken plans, so he wants evidence that someone has already absorbed a major setback and kept moving.
Cyberstarts seeded Fireblocks with $3 million in June 2018. Soon afterward, six of its first 10 design partners went bankrupt during the crypto winter, forcing Michael Chodorow to rethink the customer and plan; Raanan connects that resilience to Chodorow’s childhood immigration to Israel, growing up without a father, and upbringing by a single mother.
Life stories are only an entry point because “what’s important in a story is not the what, it’s the why.” Raanan probes job choices, relationships, motivations, and perceived superpowers; intriguingly, the best founders may be unaware of their real gift because it feels completely natural to them.

4. Sunrise starts with pain and postpones code

Sunrise grew from Raanan’s 1997 startup mistake. His team at Perfecto Technology built what he describes as the first working CAPTCHA prototype but found no business model, then pivoted into web-application firewalls; the company was later renamed Sanctum, merged with Watchfire, and was acquired by IBM.
The reversal became foundational: do not invent technology and then search for a customer. First identify painful problems, find buyers demonstrably willing to spend, and only then build something “mere mortals can use.”
Teams interview dozens of CISOs, but avoid merely asking for their “biggest pain,” because the answer changes weekly. Better prompts include “Who’s the vendor you hate the most?” More importantly, “words are cheap”: founders inspect budgets, open-source experiments, consultant-built workarounds, and vendor conversations for evidence of action.
Cyberstarts tells customers that each new team will spend about $100 million on engineering over three years and asks which single problem they would want $10 million of that effort to solve off-balance-sheet. After choosing from dozens of interviews, founders run another round to design the solution and may spend six months before coding. Later simulations assume failed evaluations or sales and work backward through product, pricing, channels, staffing, and headquarters.

5. Capital is fuel, but price can weaken alignment

Raanan’s financing principle is that “important companies are typically not cheap.” Once the product and team are right, a CEO’s priority is enough capital to hire engineering and sales capacity; valuation must support that raise without forcing founders to surrender 50% in a Series A or B.
Cyberstarts’ seed pricing changed little over seven years: first-time founders commonly received $15 million to $20 million post-money valuations, while repeat founders started around $40 million to $50 million and could climb much higher.
His nuance is deliberately conditional. Raanan has passed when pricing “went crazy,” since an investor holding only 5% may not treat the startup as a portfolio priority; yet Wiz reinforced that elite companies are expensive from seed onward. “A high priced company is actually a good thing, not a bad thing, if justified.”
Sunrise could transfer wherever buyers are relatively uniform—potentially fintech, gaming, or medical devices—but Raanan also requires a healthy acquisition ecosystem because not every company reaches public markets. Cyberstarts sold five portfolio companies for a cumulative $2.5 billion during the preceding 12 months.

6. Wiz shows why urgency beats importance

Wiz’s preparation began eight years before its founding. Its four founders built Adallom, sold it to Microsoft in 2015 after three years, and spent five more years building Microsoft’s cloud-security business to roughly $1 billion in revenue; by their 2020 departure, Raanan considered them the world’s most experienced cloud-security team.
Their first company concept, Beyond Networks, addressed secure access for satellite offices. Customers found the problem important but not urgent, and urgency is decisive for a startup because it accelerates engagement, learning, iteration, and revenue. The team pivoted into familiar cloud security and renamed the company Wiz.
Raanan’s four-persona test separates whoever feels the pain, owns the budget, holds purchasing authority, and uses the product. Four different people mean “go and pivot”; two can support a strong company, while three is borderline. Wiz had a “mega hit”: all four roles mapped to the CISO.
With AWS credentials, Wiz could reveal customer-specific value during the first prospect call. It then rejected product-led growth for high-priced enterprise sales, moving from product construction to approximately $1 million, $2 million, $8 million, and $25 million of ARR across its first four selling quarters.

7. Wiz’s scale did not make Raanan satisfied

Wiz compounded product-market fit with unusually productive engineering. Co-founder Roi Reznik sought people whose hobby was writing code—those who would use a quiet weekend hour to program. Raanan’s arithmetic: 100 such developers can resemble 300 or 400 excellent engineers, because a great developer is probably 10 times better than a good one.
Raanan describes Wiz as the fastest unicorn and fastest company to reach $100 million and $500 million, with a probable path to $1 billion ARR that year. Yet even a conditional Google transaction valuing it at $32 billion after five years did not satisfy him: “I wish I could…A/B test that decision again.”
He predicts someone will break that record within 10 years, potentially in cybersecurity, because the ingredients remain: large markets, fast-changing pain, engineering and commercial talent, experienced investors, and luck. He does not forecast the category—security follows cloud, mobile, IoT, autonomous vehicles, AI, or whatever emerges next.
That hunger sits beside profound loss. Before Cyberstarts, Raanan’s 19-year-old daughter died; he learned to “live with the pain, not fight the pain,” while working with younger people on innovation and real-world problems helped keep him younger and healthier. His investor counsel follows: find work that is “a life project.” His son later supplied his most meaningful affirmation, telling him, after reflection, “I’m his best friend.”

AI, autonomy, and the future of naval warfare with Captain Jon Haase, United States Navy ARC Prize Version 2 Launch Video! [Francois Chollet, Mike Knoop]