About
Co-Founder & GP at Blue Moon, an AI-native seed fund backing exceptional B2B founders…
Articles by Benjamin
-
🚀 If you were starting a seed fund today, would you…
🚀 If you were starting a seed fund today, would you…
1️⃣ Copy the existing playbook and compete head-to-head with established brands? 2️⃣ Or design a differentiated model —…
15 Comments -
Keep calm and manage the downside (my investment approach)
Keep calm and manage the downside (my investment approach)
Hello, my dear friends and network! I am excited to share with you what I've learned over the years of being an…
34 Comments
Activity
6K followers
-
Benjamin Orthlieb shared thisWatsi is the best place I know to have true impact on thousands of people around the world.Benjamin Orthlieb shared this**Thank you for the incredible response to our Community & Engagement Lead opening. Applications are now closed.** Are you the next member of #TeamWatsi? More than 34,000 people across 36 countries have received the healthcare they needed because strangers around the world chose to help. A child back in school after surgery. A parent able to care for their children again. A farmer back in the fields alongside his father. Every one of them is real, has a story, and is verified. And, behind each is a community of people who made care happen. We're looking for the person who will grow that community. 🌱 Learn more and help us spread the word 📢 https://lnkd.in/gzCSbWjqWatsi is hiring a Community & Engagement Lead (Remote)Watsi is hiring a Community & Engagement Lead (Remote)
-
Benjamin Orthlieb shared thisCouldn’t agree more!Benjamin Orthlieb shared thisseed investing at scale is like 97% operational, 3% strategic. almost all great investors are the opposite: very strategic, not operational. so while seed investing has the highest annualized returns of any stage in venture, it is not scalable for most investors. strategic investors move up the stack. the best seed at scale investors (like Flex Capital, Box, etc.) are the ones that are truly operational.
-
Benjamin Orthlieb shared thisAbout to start the student discussion of the HBS case about Blue Moon! Excited to hear the students ideas and questions
-
Benjamin Orthlieb shared thisI'm speaking at the Virtual DDVC Summit 2026! I'll be joining 30+ speakers from Accel, Atomico, Balderton, BlackRock, NEA and more for 3 days of practical sessions on how leading investment firms are using AI, data, and automation. My session: "Network Is the New Data: How AI-Powered Relationship Intelligence Drives Alpha" is taking place Day 1, March 23 at 7:30pm CET If you work in venture and want to know how some VC's use tech today, not theory, I'm sure you'll find it interesting! Thank you to the Data Driven VC team and Dr. Andre Retterath for having me on the panel!
-
Benjamin Orthlieb shared thisHigh concentration of ex Linkedin Biz Ops (C.2014) speakers at the Stanford VC Bootcamp yesterday, hanging out with Lutz Finger!
-
Benjamin Orthlieb posted thisThe era of the 10-tab SaaS stack is quietly ending. I recently had a fascinating conversation a serial founder who's now building "the SaaS to end all SaaS." His insight was simple but striking: 💡 When your CRM, task management, support, accounting, and AI agents all live in one system, something magical happens. It's not just convenience. It's one pool of data. And that changes everything. ‼️ → Close a deal in your CRM? It flows straight into accounting and auto-generates tasks. → Want an AI agent to run your site reliability engineering? It already knows your full context, no integration headaches. → Need to send an intro? The blurb writes itself because the system already knows both parties. We've lived this at Blue Moon. When we stopped stitching tools together and built everything around a single data layer: AI screening, deal flow, company analysis, intros, follow-ups. The leverage became real. Last year we made 278 founder-investor intros, almost entirely on autopilot. Not because we have a big team. Because the system knows the full context, reads my emails and thus generates the next step automagically. For years we've accepted "best of breed" as gospel. Pick the best CRM. The best project tool. The best support platform. Stitch them together with Zapier and prayers. But the hidden cost was never the subscriptions. It was the fragmented data. AI has now made fully integrated systems not just possible, but genuinely easy to build. One person. A few months. Now, will enterprise buyers with strict security requirements, compliance mandates, and dedicated IT teams still lean on proven, robust point solutions? Probably yes, and fairly so. But for SMBs? A fully integrated, AI-native suite that handles 90% of your workflows, bought from one place, might just be the winning ticket. ❓ The question I'm sitting with: how many of the tools in the SMB and Enterprise stack exist only because integration used to be hard? Maybe it doesn't have to be anymore.
-
Benjamin Orthlieb shared thisThink you can spot the next unicorn? Can you beat an algorithm? The HBS students get 90 minutes to debate this. You get the comment box below! 😀 Next week, we're heading to Harvard Business School to discuss a case we've been working on: Blue Moon has reimagined the entire VC workflow, from sourcing and screening to winning and portfolio management, using AI. Here's the case setup: Blue Moon built a suite of AI associates to do what most VCs do manually. Screen thousands of founders. Write personalized cold emails. Prep for meetings. Assess personality traits. The numbers are hard to argue with: 70% market coverage vs. 30% for traditional VCs, a 50% seed-to-Series A graduation rate (vs. 15% market average), and a TVPI of 4.4x five years in, beating 99.9% of their vintage peers! But the hard questions are where it gets interesting, and that's exactly what the students will be wrestling with: 🔹 Can you build conviction in a founder through an algorithm? Blue Moon assesses founders "outside in" automatically, then probes what drives a founder's resistance to pain in live meetings. Is that insight… or overreach? 🔹 Compete or collaborate? They share deal flow and AI analysis with top-tier VCs. Does that build their moat, or slowly give it away? 🔹 Should they productize? Their AI edge is real. Should they sell it, or would that kill the very advantage that defines them? 🔹 Who trusts a machine-led VC? Convincing LPs to back an AI-native fund that marries technology (billions of tokens, hundreds of thousands of lines of code) with deep personal conversations, without a traditional VC pedigree, is no small pitch. How do you sell that story? Would you back a fund where AI drives deal selection, or does that cross a line for you? Drop your view in the comments. The HBS students will be debating the same thing. 👇 #VentureCapital #AI #HBS #BlueMoon
-
Benjamin Orthlieb posted thisWe met over 50 seed stage B2B companies since Jan 1 - here are the trends that are emerging from talking to the Founders building "at the edge of the AI world". 1- compliance is turning into the product surface, 2- agents are graduating from chat to real work, and 3- supply chains are being rewired for resilience and speed. Today I'm covering trend #1: The Compliance Bottleneck Nobody Planned For The gating factor for AI isn��t accuracy; it’s governance. Teams that win are architecting for compliance up front and letting that unlock budget. Dexter’s Source‑to‑Pay overlay is a good example: it automates procurement on top of SAP/Ariba/Coupa while keeping data in the EU and enforcing identity checks time‑to‑value without a rip‑and‑replace. In finance, Interchange blends a developer platform with a regulated clearing license and direct U.S. markets connectivity, offering institutions an API surface that already fits within regulatory boundaries. Healthcare startups are taking the same tack. Reclaimly automates patient balance collections with multi‑channel outreach while integrating with EHRs under HIPAA and BAAs. Agentman’s proprietary runtime and no‑code composer target administrative workflows in similarly regulated environments, with a production eligibility agent delivered in six weeks. Even creative operations are compliance‑first: ZebraTruthAI automates long‑form ad production and bakes in guideline checks and frame‑level consistency to avoid rejections. Under the hood, servescale.ai packages on‑prem/VPC inference with a governance layer and energy‑per‑request metrics, acknowledging that sovereign deployment isn’t a niche, it’s a requirement. And from the bottom‑up, Comp AI’s open‑source agents automate evidence collection and policy management across dozens of frameworks. The shift is subtle but consequential: compliance isn’t an afterthought or a procurement hurdle; it’s a product decision that determines whether AI reaches production.
-
Benjamin Orthlieb shared thisSigned and Sealed! HBS now has a case about Blue Moon and the use of AI Agents! We're going there next month for the first class discussion of the case. Feels a bit unreal but should be really fun!
Benjamin Orthlieb commented on a post
1mo
Early on a Founder said “you have me a quick no - that’s the second best answer”. I loved that expression!
Benjamin Orthlieb commented on a post
1mo
Congratulations Christina Qi and the whole Databento team! What a journey so far!
-
Benjamin Orthlieb liked thisBenjamin Orthlieb liked thisCursor sold for $60B; OpenAI cut them off. But for Vertical AI, there's a more interesting story here... Composer *worked*. Frontier lab lifetime value leans heavily on one assumption: Build on Claude / GPT, pay Anthropic / OpenAI forever. Labs capture a perpetual double-digit margin from apps. Just like with cloud infra before it, the labs become a permanent utility. Cursor's story calls that paradigm into question. Their proprietary model, Composer, was built on Kimi K2.5, an open-weight. And it hit #3 for coding models on AACAI, outstripped only by GPT 5.5 and Opus 4.7, which are ~10–60x the cost per task. Giant pre-training budgets are no longer unshakable moats. So why did Cursor sell? Well first, $60B — the largest acquisition of a VC-backed startup ever — isn't anything to sniff at, esp for a 4-year old GM-breakeven startup. Even with a $50B-valuation round in talks with a16z and Thrive, public stock is a lot closer to liquidity. Cursor faces issues that made the independent road ahead uniquely hard: 1️⃣ A core value prop of Cursor was the ability to choose your model. This made forcing Composer near-impossible, and hence structurally reliant on retail prices for Anthropic / OpenAI models. An issue likely compounded by the fact, as a result, Cursor's harness couldn't be optimized to one model. 2️⃣ They were competing in an area of core competency for the labs: coding. Frontier harnesses — Codex and Claude Code — had been steadily eating Cursor's market share. And with every %, are gathering more data to improve. 3️⃣ Coding is uniquely well-suited to RL at scale. It's heavy on user feedback & you can simulate infinitely — it either compiles or doesn't. Labs have tons of dough for this kind of post-training. Meaning Cursor would have to keep paying to keep up. In most domains, success isn't this widely verifiable, so you need more than just $. That said, by most measures, Composer was a success. In fact, it seems to have played a critical role in tipping Cursor GM-positive after its May launch. Perhaps, in the coding space, it was just too little too late... at least to feel confident turning down a $60B offer... not to mention access to huge, wholesale compute. But there was a real path forward we can't easily dismiss. TLDR: Cursor took the model layer proprietary in perhaps the labs' best domain, on open-weight infrastructure, achieving near-frontier parity, without frontier pre-training. So what happens in domains where data is harder to access and verify? Increasing Vertical AI independence from frontier labs isn't a question of if. It's a question of when.
-
Benjamin Orthlieb liked thisBenjamin Orthlieb liked thisIn this environment, it's easy to forget that most VC liquidity doesn't come from IPOs. In 2025, for example, the median venture exit was a ~$500M acquisition. Let's dig a bit deeper. Which themes & business models drive those outcomes? Last year, of $131B app-layer proceeds across 158 exits, 56% were vertical. What stands out in that cohort: → >15 Vertical IPOs, from Wealthfront to Hinge Health → Healthcare was the most active vertical with 43 exits → Top 10 exits made up 60%+ of total value → Below $1B, PE's share of VC-backed outcomes grew massively → And here's who's most active into the next wave: General Catalyst (24 leads, $1.8B) and a16z (22 leads, $1.9B), plus new entrants like BlackRock, Mubadala, QIA What about geography? And while SF & NYC are increasingly dominant in AI-native formation, the rest of the US has played no small part: the Midwest had the highest vertical share of any region (59.6% of deals), while the West coast had the lowest (41.3%) — though that's more a story of booming AI infra than anything else. We don't expect the next waves of venture exits to mirror the last. Too much is changing, too fast. Secondaries are exploding and Vertical AI M&A has barley begun. But if history is any guide, we can say one thing: The future of VC liquidity will hinge on much more than the trillion-dollar IPO. More from our 2026 Vertical report below.
-
Benjamin Orthlieb liked thisBenjamin Orthlieb liked thisHere's how US companies are allocating dollars within the OpenAI model family. As of last week, 73% of the spend is going to GPT 5.6, with the bulk of that spent on GPT Sol. But the story to me here is just how fast dollars get cannibalized within a single lab's family of models. There are very few businesses that would spend billions to create something like GPT 5.5 and then be excited to see it shoved aside by a better product they themselves launch...in just weeks! This hints at something more fundamental, which is the pace of AI as a strain on business. For the model makers, you're only as good as your most recent model and that better have been weeks (not months) ago. For businesses, you're consistently badgered by fear that a different model, a cheaper one, a more specialized one, whatever, could 10x your app. For consumers....well, this is kinda less impactful for consumers. You hope the model makers abstract away this decision while you benefit from ever-increasing intelligence. How quickly to do you hop to the frontier model in the family?
-
Benjamin Orthlieb liked thisBenjamin Orthlieb liked thisAI is creating a two-sided story for M&A. It's fueling deal activity in areas like energy infrastructure and semiconductors, while also making future market positions and valuations harder to predict, particularly in technology. I'm also watching how heightened attention on a few high-profile transactions is shaping perceptions, though I'm careful not to read too much into any single deal and extrapolate it across the broader market. I sat down with Kristen Scholer from NYSE Live to unpack these dynamics and where I see the dealmaking environment heading.
-
Benjamin Orthlieb reacted on thisBenjamin Orthlieb reacted on thisA very British goodbye courtesy of The King's Guard. RIP Dolly.
-
Benjamin Orthlieb liked thisBenjamin Orthlieb liked thisThe same lead can convert three to five times better depending on who follows up. I saw that pattern over and over at Zillow, and it is the reason we started HouseWhisper. I got into it with Ryan Rodenbeck on the #RealtyHack podcast. A few things we covered: 🟡𝐓𝐡𝐞 𝐠𝐚𝐩 𝐢𝐬 𝐧𝐨𝐭 𝐥𝐞𝐚𝐝𝐬, 𝐢𝐭 𝐢𝐬 𝐟𝐨𝐥𝐥𝐨𝐰 𝐮𝐩. Every agent has access to the same portals. What separates top producers is the process after the lead comes in, sustained over months, sometimes years. 🟡𝐀𝐠𝐞𝐧𝐭𝐬 𝐬𝐡𝐨𝐮��𝐝 𝐛𝐞 𝐚𝐩𝐩𝐫𝐨𝐯𝐞𝐫𝐬, 𝐧𝐨𝐭 𝐝𝐚𝐭𝐚 𝐞𝐧𝐭𝐫𝐲. Your CRM should already be updated when you get back to the car. You review and approve, you do not type. 🟡𝐍𝐨 𝐨𝐧𝐞 𝐧𝐞𝐞𝐝𝐬 𝐚𝐧𝐨𝐭𝐡𝐞𝐫 𝐝𝐚𝐬𝐡𝐛𝐨𝐚𝐫𝐝. Tool fatigue is the reason great software goes unused. HouseWhisper works over calls and texts because there is nothing new to learn. 🟡𝐓𝐰𝐨 𝐠𝐨𝐨𝐝 𝐭𝐚𝐬𝐤𝐬 𝐛𝐞𝐚𝐭 𝐟𝐢𝐟𝐭𝐲. An assistant that hands you a wall of to-dos is just noise. The hard part is picking the one or two that actually move a deal forward. We also got into what happens to proptech as AI lowers the barrier to building software. My take: the pie is growing faster than it is getting split, and the gap between a cool demo and something thousands of agents rely on every day is still enormous. Thanks to Ryan Rodenbeck for a sharp conversation. #AIforRealEstate #HouseWhisper #RealtyHack
-
Benjamin Orthlieb liked thisBenjamin Orthlieb liked thisHappy birthday to my one and only co-founder Ayan Barua. I can always count on you to bring the energy, and hit the target. 🎯
-
Benjamin Orthlieb liked thisSuper excited to open this next chapter for Honeycomb Insurance! go LRO!Benjamin Orthlieb liked thisBreaking news: Honeycomb Insurance now covers office and retail properties after launching our Lessor's Risk Only (LRO) program 🏢 🐝 For years we've brought tech-enabled quoting to landlords and condo associations. Now we're bringing it to retail and office properties. Our new Lessor's Risk Only (LRO) program is officially live in Minnesota, Arizona, Colorado and North Carolina, with more states rolling out over the coming months. LRO covers property owners who lease space to tenants, and we're launching with capacity up to $25M per location. What's in appetite: • Well-maintained office and retail buildings with everyday daytime tenants • Professional offices, salons, boutiques, and the classic strip mall footprint • Restaurants! 🔗 Full story: https://lnkd.in/ggFdrABX #HoneycombInsurance #LRO #CommercialRealEstate #InsurTech #CommercialInsurance #LessorsRisk #InsuranceAgents #PropertyInsurance #RealEstateInsurance
-
Benjamin Orthlieb liked this🎉 🎉 Awesome deal for Checkr, Inc. and Truv! Wilson Sonsini Goodrich & Rosati really enjoyed working with the whole Checkr, Inc. team, including Toni Llorens, Jennifer Yeh, Raul Moreno, Keith Mariano, and (for most of it 😉) Michelle Pham. We're very excited to see Checkr, Inc. move into the Government space, exciting times! 🎉 🎉 #mergersandacquistions #techlaw #artificialintelligence #government #mortgagesBenjamin Orthlieb liked thisWe're excited to announce that we have acquired Truv, a leader in consumer-permissioned income, employment, and asset verification. Truv's network already reaches 96% of the U.S. workforce and serves mortgage lenders nationwide, relationships that take years to build. This acquisition marks our entry into Government, strengthens our Mortgage offering, and advances our vision of becoming the most comprehensive, AI-powered verification platform for high-stakes decisions. With identity, background, income, employment, and asset verification together in one place, customers can make faster, more confident decisions for the moments that matter most, while consumers benefit from increased transparency and accuracy. Welcome to the team, Truv! 🤝 Learn more → https://lnkd.in/gtpAv8rw
Experience & Education
Licenses & Certifications
Publications
Other similar profiles
-
Gautam Tambay
Gautam Tambay
At Springboard, we're building the future of education, with the help of an incredibly talented, mission-driven team. Join us! We'll transform lives together.<br><br>Previously in my career, I've been a startup utility player, strategy consultant, venture and private equity investor and startup advisor. I'm passionate about startups, education, the outdoors, food, and India.
-
Haggai Klorman-Eraqi
Haggai Klorman-Eraqi
Explore more posts
-
Ritu Bhatt
Artificial intelligence is rapidly reshaping the global research and innovation landscape. Over the past decade, China has transformed itself from a fast-growing participant into the world’s largest producer of AI research by publication volume and citations, while the United States continues to lead in frontier AI models, computing infrastructure, private investment and commercialization. India has also made significant progress, particularly in AI publications and its expanding pool of technical talent, but has yet to develop the same depth of research institutions, funding, computing capacity and university–industry linkages. This infographic examines how China overtook the U.S. in AI research output, why the U.S. remains dominant in frontier AI, and why India has not yet followed China’s trajectory. It highlights the structural factors that determine whether research translates into patents, technologies, companies and global technological leadership. #ArtificialIntelligence #AIResearch #AIInnovation #China #UnitedStates #India #AIIndia #Technology #DeepTech #ResearchAndDevelopment #AIResearch #InnovationEcosystem #TechPolicy #FutureOfAI #AILeadership #GlobalInnovation #Research #HigherEducation #STEM #DigitalIndia #IndiaAI #ChinaAI #USAI #FrontierAI #AIInfrastructure #DeepTechIndia #TechnologyPolicy #InnovationStrategy #FutureOfTechnology #AI2026
1 Comment -
Jeff Johnson
The hardest companies to build are often the ones the world needs most. Resilience Tech—spanning hardware, software, deep tech, and infrastructure—demands more than just great ideas. The real test isn’t proving the technology works—it’s building it into a company that can scale and last. At B Capital, Karly Wentz and I distilled what we’ve learned into a 10-factor framework for growth investing in this space. It’s how we separate promising pilots from companies that can scale into lasting platforms. Our conviction: the companies that matter most over the next decade will be those that fuse breakthrough science with business fundamentals—and deliver measurable resilience at scale. Read the full article here: https://lnkd.in/gv8jaMPK
11 Comments -
Evan Nisselson
Jazzed for this keynote by Jeff Erhardt at our 12th annual LDV Capital Vision Summit: The Materials Innovation Gap: Why AI Predictions Aren't Enough — And What It Will Take to Transform a $6 Trillion Industry RSVP: https://lnkd.in/eeUHN_m AI is transforming industry after industry, and generating enormous excitement in the world of materials science. Machine learning models can now predict millions of new material candidates with unprecedented speed and accuracy. At the same time, programmable material classes, including those recognized by the 2025 The Nobel Prize in Chemistry, could unlock breakthroughs in some of the most pressing challenges facing civilization, from environmental remediation to the energy transition. But their impact is constrained by something prediction alone cannot resolve: new materials must be physically produced, scaled, integrated, and qualified within the larger systems into which they are deployed, a journey that remains extraordinarily long, expensive, and unpredictable. Jeff will argue that closing this materials innovation gap requires treating AI and physical experimentation as equal partners, and building a new model of collaborative development that connects materials innovators and industrial partners far earlier in the discovery process.
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content