Andrei Craciunescu

At #ITCVegas, I was about the future of insurance. We covered: • The story behind RiskCube • How founders will buy insurance in the AI era • What thriving in 2030 looks like for insurers Excited to share more soon!

@RampLabs That’s amazing!

Today, we're launching Ramp Agent Cards. There's been no safe way for agents to spend money, until now. Ramp Agent Cards give agents the ability to spend, governed with real spend limits, merchant controls, and full visibility into every transaction.

@SenSanders We are an AI insurance broker in CA. We take data privacy very seriously and find ways to remain AI-native while also protecting our customers' data. It's possible to have both.

I spoke to Anthropic’s AI agent Claude about AI collecting massive amounts of personal data and how that information is being used to violate our privacy rights. What an AI agent says about the dangers of AI is shocking and should wake us up.

@daltonc @paultoo Great episode!

New Dalton + Michael episode released: "How to get unique startup ideas" One of the biggest problems founders and (builders in general) have right now is not "how do I build my MVP?" but instead "what idea should I be working on?" It feels like all of the good ideas are taken, and the competition for those ideas is ruthless. In this episode, we discuss some of the starting conditions for coming up with unique startup ideas, and why it can feel so hard to come up with something that is both: 1) something people want and 2) original. We discuss some tactics you consider trying that are more likely to yield differentiated startup ideas vs derivations of whatever The Current Thing is. 🔗 below

Some of the founders building autopilots across these verticals 👆@ryanjdaniels @mitch_troy @jakeserval @BrendanFoody @davepaffenholz @andrei_serban @justindross @ChKashifAli @nicckopp @StephenHedlund @askvladi Who should we add?

@JulienBek @ryanjdaniels @mitch_troy @jakeserval @BrendanFoody @davepaffenholz @andrei_serban @justindross @ChKashifAli @nicckopp @StephenHedlund @askvladi We are building a YC-backed AI-native insurance brokerages! Happy to chat! riskcube.com

@saranormous The insurance idea was the best!

Embed AI Grant is the best deal for elite AI startups, 10 teams, 2x a year: - $250K in cash (uncapped SAFE) - $1M of credits - retreat with founders, builders/research leaders - community working at the frontier - investor demos with every top VC firm embed.conviction.com

Last 5 AI-native startup ideas from us @Conviction before Embed apps close (robotics cert, red-teaming chips, materializing institutional memory, physics sim, and getting assurance on abundant generated code)

@saranormous @conviction I like the insurance idea! This is going in the right direction.

Deadline for @conviction Embed ($250K grant, compute and tools credits, community and mentorship program for the 10 best AI startups) extended to Monday 11:59PM PT. As is tradition, here are some of startup ideas we've been pondering and cooking -- apply if you are too!

@jefielding Deep tech will always be relevant since it’s more defensible.

Half the VCs I know are changing their focus areas / investment thesis right now. Feels like a moment of deep reflection - or panic.

@PeterJ_Walker Any information about round sizes?

Fresh data from Silicon Valley Bank on ARR at time of fundraise. Chart is mine but follows their original work which you should read. Google SVB state of markets H1 2026. Seed - Median: $0.22M - 75th pct: $0.77M - 90th pct: $3.6M Massively wide ranges across early stage

@mad_ins_agent Where are you on the journey?

I’ve been working on CPCU for like 8 years. Why am I doing this?

New YC homepage design just launched! 👀 The last few years have been a return to our roots at YC, so we wanted to capture that with our new site, and show what it's really like to go through the batch. There's also a bunch of fun easter eggs throughout – thread below!

@eve_bouff @ycombinator @aaron_epstein @garrytan Truly amazing!

Designing the new @ycombinator website with @aaron_epstein and @garrytan was a dream come true. I viscerally believe YC is one of the most iconic and impactful brands in the world, and having the chance to help shape how it shows up in the world was an honor.

@ferzhao @ycombinator @snowmaker Congrats!

And now we’re a part of @ycombinator W26 batch! Thanks @snowmaker for taking a chance on us. Will be posting more updates here

Got rejected from every accelerator but YC Here's my story

@garrytan @snigdhasur Always consider buying D&O insurance! It should be the first insurance a founder should buy to protect their personal assets.

Namecheap is suing a customer and startup founder, naming the founder individually This is no way to run a business

Thanks for sharing! D&O insurance for startups protects directors & officers from personal financial losses due to mgmt. decision lawsuits, incl. legal fees, settlements, & judgments for contract disputes or IP issues. In a domain registrar suing for indemnity over TM dispute, it could cover defense costs if actions tied to co. duties, tho depends on policy details.

A cautionary tale for fellow startup founders: @namecheap just sued me personally 🧵

With rules like this, who would invest in a German startup? Not me.

@bryan_johnson @grok What should be used instead?

If you have a non-stick cookware. Get rid of it. New study: a 169% increase in the risk of fatty liver disease in adolescents is linked to PFOA, a forever chemical used in non-stick cookware, with every doubling of exposure. I've been meticulously measuring these toxin levels in myself and drinking water. Significant study findings: + PFOA showed the highest risk increase for fatty liver disease in adolescents; each doubling of PFOA plasma concentration was associated with a 169% increased odds of fatty liver disease. + In one cohort, each doubling of PFHpA was associated with a 73% increased risk of fatty liver disease in adolescents. + Age mediated the risk posed by PFOA in adolescents; each year increase in age added a 45% increased odds of fatty liver disease for every doubling of PFOA plasma concentration. + The high-risk PNPLA3 genotype significantly increased the risk of fatty liver disease with PFAS exposure in adolescents, particularly for PFHxS. Each doubling in PFHxS concentration was associated with a 552% increase in fatty liver disease odds in older adolescents carrying the PNPLA3 GG high-risk phenotype. + No basic association with increased risk was observed with increasing PFAS concentrations in young adults. + Smoking in young adults turned several PFAS, including PFDA, PFHpS, and PFNA, into significant risk factors for metabolic liver disease. Added Context: + MASLD definition: Having >5.5% liver fat in addition to at least one metabolic disease criterion (high BMI, fasting glucose, blood pressure, triglycerides, and/or low HDL). + PNPLA3: A gene involved in fat and cholesterol handling and storage in the liver and adipose tissue. The GG phenotype predisposes its carriers to an increased risk of developing fatty liver and metabolic (non-alcoholic) liver disease. Significance: Known as "forever chemicals," PFAS persist for years or decades in the human body (PFOA has a half-life of 1.5–5 years) and for decades to centuries in the ecosystem (PFOA has a half-life of 92 years in environmental water). PFAS are widely used as insulating and non-stick materials in non-stick cookware, clothing, electronics, and many other household and industrial applications. PFAS have long been known to cause metabolic and endocrine disruptions, affecting metabolism, insulin function, and liver fat metabolism. Adolescence and early adulthood are particularly sensitive phases due to hormonal changes around puberty and related rapid growth, making this age group especially vulnerable to the endocrine, hormonal, and metabolic damage caused by PFAS. This study successfully identified adolescents as an especially vulnerable risk group to specific types of PFAS (PFOA and PFHpA). Given that PFOA is used in non-stick ware and other food containers, this highlights the importance of avoiding these items in households with children, especially around and after the age of puberty. Furthermore, the study uncovered the exacerbation of genetic risk by PFAS, indicating that PNPLA3 GG carriers should be especially aware of the risk from PFHxS. This chemical is commonly used in firefighting foams, water-repellant clothing and polishes, as well as various electronics. Young adults showed no basic association between PFAS and liver disease risk, possibly indicating a special sensitivity before and around puberty. However, smoking altered this relationship, making young-adult smokers more susceptible to increased liver disease risk with PFAS exposure. Limitations While the study offers several actionable insights, more research is needed to determine if the findings hold for non-Hispanic adolescent populations (as the SOLAR adolescent cohort was solely of Hispanic background). Additionally, the somewhat small sample size limits the study's statistical power. Finally, the single-time point PFAS exposure assessment and the cross-sectional nature of the study make it impossible to assess cumulative exposures or draw any causative conclusions, limiting the insights to observational findings.

@WillieWarbucks @Bradleyflowers There are many examples that tried to do the same but failed. In my opinion, the way forward is to empower brokers, not replace them.

@andreicscu @Bradleyflowers AI can be a super a super useful tool. I use it a ton. But it doesn’t replace a risk manager. The space they’re trying to play in requires one.

Funny how there are 1.43 million real estate brokers in the US as of 2025

@WillieWarbucks @Bradleyflowers I agree, I see many AI brokers nowadays and nobody really understands risk management for insurance. This is what AI can't do. They just benchmark, from what I see.

@Bradleyflowers Ya, these guys have a slick platform but zero insurance knowledge and are going after a client base where risk assessment really matters. Basically trying to do what many have done for micro business but instead will really harm someone’s program when real assets are in play.

@t_blom @jonas Great to see @ycombinator in Munich!

Y Combinator kommt nach München! We'll be joined by my Monzo cofounder @jonas who is a CDTM alum. Apply here: events.ycombinator.com/yc-munich-2026