Matt

Disclaimer: All content I post or comment on this platform is for informational and entertainment purposes only. Nothing here should be considered legal, financial, or professional advice. Always consult a qualified professional for specific advice. Long LMND, TSLA, BlockChain and many others. I only trade stocks and have no short positions. I invest in stocks I feel are undervalued by human nature, leveraging anchor theory to determine when to buy and sell primarily. I’m ENTP and love nothing more than a good debate.

@ArranGray9 100% they can build models without complying with data and privacy rules

@sanderdaehlie Talks with her brother about insurance generally, more about insurance from a users perspective - not really anything new from a lemonade platform or business perspective - but of background on why they went full stack provider and not a technology platform

@Anchor_Invest What’s the TLDW?

youtu.be/nWnOQbLTPEQ?si… Interview with Maya

@jrouldz Pretty light on LMND there 👀👀 😂

Updated breakdown of my main portfolio with accompanying pie chart. Big pie chart guy $IONQ 12.9% $PLTR 8.5% $AMZN 7.1% $GOOG 5.2% $RGTI 4.9% $META 4.8% $RKLB 4.4% $QBTS 4.2% $XYZ 3.8% $ARKG 3.4% $SOFI 2.8% $LMND 2.8% $SHOP 2.7% $MSFT 2.5% $CRWD 2.4% $NET 2.2% $TSLA 2.2% $DOCS 2.1% $BLDE 2.1% $UBER 2.0% Options 2% Other 15% This portfolio makes up about 1/3 of my net worth

LemonApe daily digest 🍋🦧 - Everyone thinks you can bolt on an LLM and match Lemonade. You can’t. Because Lemonade’s system can act - and that’s what most people are missing!! let me explain... Since LLMs went mainstream, the assumption I keep hearing is this: "Just add an LLM to a traditional insurer’s stack and you’ve basically got Lemonade." No! And I want to explain - in plain language - why that’s just not how this works. Yes, Lemonade uses LLMs. Maya and Jim automate the customer and claims experience. But their real power comes from the system underneath - the part you don’t see. That includes the Forensic Graph that understands how behaviours are connected. The Customer Cortex that coordinates decisions. And real-time feedback loops across pricing, risk, fraud, and ops. It’s not just an LLM. It’s a totally unique, native AI - a living system. Here’s the important bit most people overlook: LLMs are great at understanding and generating language. They power chatbots, responses, document summaries, and CX automation. But they don’t understand how things or people are related. They don’t detect patterns in behaviour. A network graph does. It maps relationships between people, devices, accounts, claims, IPs, and timing. It learns what normal looks like, and flags when something doesn’t add up - even when nothing looks obviously wrong. That’s how you catch the early signals of fraud, exposure, or coordinated risk. LLMs can talk. Graphs can think. Lemonade built a system that can do both - but more so, can act! - this is what's unique and missed! Now, I know I’ll get responses like: “But you can build graph intelligence into a legacy insurer.” Yes, you can - companies like Palantir (PLTR) offer tools like Foundry that can absolutely power graph analysis, even for fraud detection in insurance. But here’s the difference: Palantir is a framework. You still have to ingest your data, clean and model it, build ontologies, connect it to systems, integrate the decisioning, train your teams to use it, and maintain it. It’s powerful tech, but it’s not plug-and-play. And more importantly, it’s not already connected to your quote, bind, claim, comms, payment, HR, or marketing flows - unless you build all of that. And that’s the HARD bit! Most insurers are dealing with 30+ years of legacy technical debt. People know they’re behind, but they underestimate how risk-averse these organisations are, how slow decision-making is, and how heavy compliance and culture weigh everything down. Early in my career, I was provisioned at a large insurer to do business process mapping. It took more than three months - four of us full-time - to map out internal systems, speak with each team, and document every process flow and system touchpoint. Even with all that detail, I can promise you - it was flawed!! And that was just to get to the point where they could even issue an RFP for a new policy admin platform. The project to replace the policy admin system took many, many years! and just as it was being deployed they were thinking about ripping it out again.... Lemonade already built that. Everything talks to everything.... seamlessly. I can't stress how unique and powerful this is. The moment the Forensic Graph spots a pattern, pricing can adjust, marketing can pause, fraud scoring changes, and approvals slow - automatically. Not in a quarterly meeting. Not via email. It just happens. Because it’s one system - not ten (or 600!) trying to talk to each other. I’ve seen this shift before. In banking, I saw rule-based fraud engines - built over decades with millions of rows of logic - get blown away in six weeks by a graph model, yet the wider interconnectivity across the business never emerged. Specifically on the Graph Network, It was fast, accurate, and self-learning. That same change is coming to insurance. Slower, because the data is thinner, but the destination is the same. It’s just a matter of time. And Lemonade’s not just using AI for customers. At InsureTech, Adina shared something that more people need to talk about - They’re using AI to help their people grow. The HR system actively requests 360 feedback, analyses strengths and gaps, and provides regular, detailed, completely unbiased advice to employees on how to improve. It’s a feedback loop for human capital. Always on. Always improving. Just like the rest of the platform. This is what Lemonade actually has: LLMs to talk Graphs to think A Cortex to act Everything - from HR to pricing - connected AI embedded everywhere, not just in the UI Systems that learn from each other, not in isolation A culture of fast-fail and quality product design An ethos of customer-first Most insurers apply AI at the edge, Lemonade built it into the core. This isn’t a chatbot sitting on top of a 30-year-old claims engine. It’s a fully integrated system (human and synthetics) that gets smarter every day. And that’s why you can’t just bolt on Foundry or an LLM and expect the same outcome. So yes - adding Foundry or a similar platform might help traditional insurers improve fraud detection, match rate to risk more dynamically, maybe even automate some doc review. But that won’t give you a system that can act. That won’t connect everything across CX, risk, ops, and marketing. And it won’t solve for the biggest opportunity left in insurance - the expense ratio. That’s where real operating leverage will come from, it's far more nuanced than just matching rate with risk… Because if you rely on third-party systems, you’ll always have to: Pay them! Wait for them Consult with them Depend on their roadmap Compete for attention with 1,000+ other clients That’s not how you build a first-party, customer-first business! It’s not a silver bullet, it's a plaster trying to stem the bleed from an artery. And it’s definitely not what Lemonade built. Lemonade is a native AI company in the truest sense of the word - a totally unique and novel approach to insurance, with a culture that pushes the boundaries of what’s possible. From the top down, they embrace risk, aren’t afraid to fail, and constantly learn and adapt. You can’t buy culture. And you can’t buy your way out of technical debt - not without starting from scratch... But that introduces the classic innovator’s dilemma: To rebuild, a company must disrupt itself from the inside out. That’s hard to pull off while keeping shareholders and employees aligned. It usually takes longer than a typical CEO’s tenure... and every board reshuffle tends to bring a new strategy - more often than not leading back to square one, or worse, spinning in circles without delivering real change. This is why you can't just add an LLM / ML model and expect to yield the same outcomes. IMO $LMND is unique, a diamond in the rough and long term will be one of the first autonomous companies generating unprecedented amounts of free cash flow and shareholders returns. If you found this useful and want more, feel free to follow for more LMND content in future - appreciate the support!

@shai_wininger @hypergrowth102 @elonmusk 👀

@hypergrowth102 .@elonmusk makes great cars, but we make better insurance

$LMND quote $200-300 cheaper than $TSLA insurance in AZ 🤯

Stock sales from SoftBank roughly 0.1% of holdings and Tim executed some options and flipped them. Tim’s share count remains NET Neutral but he pocketed some cash. Tim’s was part of a pre-arranged sale plan organised months ago, not based on anything happening right now.

“Lemonade paid my claim in 3 seconds” - and why you can’t just bolt on an LLM and expect the same results Not 3 days. Not 3 hours. Literally 3 seconds. Here’s how they’re rethinking insurance payouts and why that matters when you’re stuck, stressed, and need emergency help fast. A few years ago, I was called into a meeting with a large insurer. Their marketing team had just launched a product called Emergency Care – simple idea: if a customer lost access to their home, they’d help with hotel costs, food, and essentials. It sounded great on paper. But there was a big problem.. They hadn’t realised banks don’t move money over weekends or bank holidays. 🤦‍♂️ Customers were signing up, and it mostly worked during the week – but even then, it still took 2 to 24 hours to get the funds to them. So while it aimed to offer immediate emergency support, in practice it rarely felt immediate. And the moment someone needed help on a Fri night, Sat, or Sun, they couldn’t actually facilitate the service they’d sold. No payments until Monday. That moment stuck with me. A perfect example of what happens when customer-facing ideas aren’t backed by execution-ready tech. And that’s where Lemonade stands out. Their AI bot Jim handles claims instantly – if fraud checks are passed and it’s a covered loss, the payout can be approved and sent within seconds. In fact, Lemonade holds the world record for the fastest insurance claim payout – just 3 seconds from claim to cash. That’s made possible through systems like Hyperwallet (via PayPal), which lets Lemonade send funds programmatically to hundreds of countries, in payment methods people actually use: •Bank transfers •Prepaid debit cards •PayPal •Venmo •Even cash pickup And some of these rails work 24/7/365. So if someone needs emergency funds on a weekend or holiday, they get it – even when banks are offline. In the past, I’ve asked insurers why they still rely on outdated claim payout methods - like mailing cheques or manual transfers. Usually tech debt and “We want to hold onto the float as long as possible to earn on it.” That always rubbed me the wrong way. It just feels… anti-customer. Lemonade flips that mentality. Fast payouts, local options, no friction. It’s a better experience – and a smarter platform strategy. By handling payouts through a single API and global infrastructure, Lemonade can scale into new markets without rebuilding processes or sourcing new providers. One platform. One integration. Global reach. And it’s not just about efficiency. It also unlocks smarter risk management. For example: instead of reimbursing a customer for car repairs, Lemonade could issue a prepaid debit card to a third-party garage. If the garage acted fraudulently, Lemonade could dispute the transaction – like a chargeback. It adds a natural layer of defence. I don’t know if Lemonade uses this approach today, or exactly which payout methods they support – but the point is, they could. The rails and logic are there. And here’s the key point: This isn’t just about having an AI. Plenty of insurers are adding LLMs to legacy stacks and calling it innovation. But if your systems can’t act, your AI is just a fancy front end! Take a common scenario: a customer submits a claim for a stolen laptop. A traditional insurer might use an LLM to explain the process – but the moment the claim needs review, approval, or payout, the LLM hits a wall. It’s not connected to the policy system, the fraud engine, or payment rails. The result? A helpful chatbot that can’t do anything. Lemonade is different. Their stack is fully integrated. Jim doesn’t just chat – he verifies coverage, checks for fraud, approves the claim, and triggers the payout. All in seconds. Without that backend, no LLM can replicate the experience. That’s how you go from AI assistant to real-world results: •Better customer outcomes •Lower fraud •Scalable infrastructure •And a meaningful edge over incumbents

@shai_wininger @Lemonade_Inc New State for Car, new State for Pet, $1bn IFP can we end out this week with 55% - 60% of all tickets automated yet??🤞LLM.ND

.@Lemonade_Inc Pet is now available in Louisiana—our 40th pet state

@jrouldz Bought more today

$IONQ One of these is not like the others 😏

@ssmith3088 There’s a lot of irony using the Churchill dog 👍

linkedin.com/posts/lemonade… Pet insurance expanding into new states

@GrindeOptions More. A lot more.

Have you ever lost $15,000+ in one day inside your stock portfolio? 👀

@unusual_whales Pete using Chat GPT to polish his top secret military texts is next level 🤦‍♂️ Long dash’s and structured paragraphs- dead give away.

BREAKING: Editor in chief Jeffrey Goldberg has released the Signal war plans text thread. See them here: unusualwhales.com/news/atlantic-…

On the train yesterday, I look over my shoulder to the left, the person next to me is not on Facebook or websites - just chatting away on Chat GPT, look to my right and the same thing on that side - I look down at my own phone - and yep…. I was only looking up because I was waiting for it to do a deep research…. Monumental shift is happening, and very fast.

MELIUS JUST SAID $GOOGL RISKS LOSING YOUNGER USERS TO OPENAI SEARCH 😳

@conneryoungBC Opex excluding marketing has been flat for 2 years…. Despite the company being far far larger now. 🚀

Someone tell me I'm not crazy looking at $LMND and if they grew their Total Revenue by 7x and their in force premium by 5x that means they're getting way better at making money off of their customers... Why does this meme stock not look like a meme stock?

@riyadosofian Did this a while back - not the latest numbers but you get the point…

$LMND 2023 Number of Customers per Employee = 1,600 2024 Number of Customers per Employee = 2,000 That's a 25% YoY increase. Legacy Insurers? Only about 300. And it was flat YoY. $LMND has some serious structural advantage right there.

@amitisinvesting LMND

What is one stock that you think does not get enough media coverage but is a legitimate company that can outperform the S&P over the next 5 years? On a flight right now & I told myself I would finally watch the Sopranos but I would much rather do DD instead 😂 I speed-watched (it’s like speed reading) severance over the weekend. Good show. Currently reading about $ASML — would love to get other names below and any thesis associated with the names!

@energy_tech_inv Sold most of my crypto when BTC was $110k and sold out the rest at like $94k, flipped it into LMND - still very bullish on BTC long term. Very, very bullish.

Considering selling a large portion of my remaining $ETH to buy more $LMND I’ve held my $ETH for a weighted average of 3.5-4 years. Not certain of $ETH’s upside potential from here or its overall ecosystem. I used to be on top of all the ecosystem’s new developments.

@SK_Invests007 And that’s why I won’t share who liked this first…..

I get it - at the same time - I think whether Management or PB or any other person, they should be safe to “like” something here without their opinions suddenly being made public. I think people here will be far more open to giving their opinions when they know this community is a safe anonymous place. 🤷‍♂️

Daniel liked a Lemonape's comment to push that gas pedal hard 😜. Jokes aside... I feel, team's confidence is growing... they have been controlling the growth, I'm sure now they are confident to accelerate the growth. Let's go. $LMND