He spent a decade buying companies for LinkedIn. Now he runs a seed fund that behaves like a fintech, with six AI agents on the payroll and a pipeline of 9,000 founders a year.
Ben Orthlieb opens the laptop at 7 a.m. and an AI named Agatha has already done the night shift. She has read through thousands of digital footprints, ranked the founders she likes best, and flagged a short list for a human to look at. By noon, a second agent named Da Vinci has read 200,000-plus pages of curated early-stage content and prepared a brief on the company. By the next afternoon the founder has a yes or a no.
This is what Blue Moon VC looks like from the inside. It is what Orthlieb left LinkedIn to build - a seed-stage fund where machines do the reading and the partners do the deciding. The original name of the firm was 2.12: two meetings, an answer by noon the next day. The new name is softer, the model is the same.
The math is the part Orthlieb would rather you focus on. Across the 2021, 2022 and 2023 vintages, Blue Moon reports a 45 to 50 percent graduation rate from seed to Series A. The market average sits around 14 percent. Even the top venture firms hover near 28. Two companies from the 2021 vintage have already crossed the unicorn line.
At the early stage, at seed, where we play, 90% of what matters is really the founders.
- Ben Orthlieb, Driving Alpha podcast
Orthlieb's pitch to LPs is unfashionably simple. Most seed funds, he argues, are built around access and brand. Without either, you are stuck with the deals nobody else wanted. His answer is to industrialize the front end of the funnel and rent the brand by always co-investing with a top-fifty fund. The data filters out the noise. The signal VC filters out the adverse selection. The partner time, what's left of it, gets spent on the founder.
Sources: Blue Moon VC, Alpha Partners' Driving Alpha podcast, Family Office Insights Q&A. Self-reported.
Graduation rate from seed to Series A, as reported by Ben Orthlieb on the Driving Alpha podcast (Alpha Partners, 2025) covering Blue Moon's 2021-2023 vintages.
Before Blue Moon, before 2.12, before the AI agents with detective names, Orthlieb did the conventional rotation of an ambitious French-trained operator landing in American finance. There was BNP Paribas. There was Barclays. There was Boston Consulting Group. There was Oracle, where he worked on the Public Cloud strategy team. And then there was LinkedIn, where he stayed for more than a decade and rose to VP and Head of Corporate Development, running mergers, acquisitions and investments. He worked closely with Reid Hoffman. He bought companies for a living.
Somewhere along the way, on the side, he started scouting for Emergence Capital - the enterprise SaaS firm best known for backing Salesforce, Veeva and Zoom. Scouting taught him something his day job could not: that the bottleneck in venture is not opinions, it is reps. The best partners see thousands of deals before they get good at picking. The worst part of the job is that you cannot get to those reps without burning through a fund or two of LP money first.
So he and Romain Serman, his co-founder, built a system that could do the reps without humans. They named the first version of the firm after its own process - two meetings, twelve the next day - and started writing small checks alongside the signal funds. The 2021 cohort produced two unicorns. The LPs noticed. Fund II got bigger. The branding got softer. Blue Moon was born.
The thing to know about Orthlieb is that he is a corporate-development person at heart, not a stage-stalking partner. He has spent his career evaluating other people's businesses through the lens of someone who actually had to integrate them afterward. The Blue Moon process is just that lens, rebuilt for seed.
Each Blue Moon agent owns one slice of the workflow. The names are part inside joke, part marketing: detectives, painters and a financial terminal. The work, less romantic, is largely retrieval, ranking, and scheduling.
The sourcing brain. Scrapes founder backgrounds, parses signal from digital footprints, and ranks roughly 9,000 teams a year into a triaged 300-400 the humans actually meet.
The analyst. A retrieval-augmented system trained on 200,000-plus pages of curated venture content. Briefs the partners before founder meetings.
The chatbot. Blue Moon claims it answers founder questions with an 80% performance advantage over a generic GPT in this domain.
The portfolio detective. Tracks performance, flags anomalies, and watches the metrics so the partners can spend their time on founder calls.
Deal flow coordination with the top fifty US funds. Routes intros, surfaces co-invest opportunities, keeps the calendar of who is looking at what.
Operations orchestrator. Manages the workflow between the other five agents and the two-plus million automated operations Blue Moon runs each year.
Seed only. North American B2B enterprise SaaS. AI/ML startups and deep technology where founder quality is the dominant variable. Nothing later, nothing consumer.
$250K to $500K. Designed to slot in alongside a larger lead and add disproportionate operational value per dollar.
Blue Moon co-invests with a "signal VC" from the top fifty firms - by policy, not preference. The reasoning is brutal and correct: without a brand, you cannot risk adverse selection.
Last year alone Blue Moon made 278 warm introductions to top VCs on behalf of its portfolio. The fund's main service to founders is access.
At the early stage, at seed, where we play, 90% of what matters is really the founders.Driving Alpha podcast, Alpha Partners, 2025
Two meetings, answer by next day at 12.The origin story of the firm's original name, 2.12
The original firm name had nothing to do with February 12 or a Manhattan address. It was a process spec: two founder meetings, an answer by noon the next day. The name described the SLA.
The agents are named after Agatha Christie, Leonardo Da Vinci, Raphael, Kojak the bald TV detective, the Bloomberg terminal, and the punctuation mark Slash. A reasonable description of how most of venture actually feels.
2.12 reported a portfolio composition roughly ten times more diverse than the seed-stage market on women and underrepresented founders. The argument: when you screen on data instead of networks, the network bias disappears.
"Benjamin Orthlieb, Blue Moon GP on B2B SaaS, Data & AI-Driven VC."
Alpha Partners' Steve Brotman sits down with Orthlieb on how a founder-first, AI-powered approach is quietly tripling market graduation rates.
"AI Native VC, Achieving 50%+ Graduation from Seed to Series A, Why Access Is the Key to Success."
Most seed investors learn the craft by losing money for a decade. Orthlieb skipped that step by spending it inside LinkedIn instead, on the buy side of a hundred conversations about whether a company was worth what someone said it was. That is the angle Blue Moon represents - not a fintech wearing a venture costume, but a corporate-development practice wearing a venture license, scaled up by software so it can be applied to thousands of companies a year instead of dozens.
The bet is that the seed market, which has always been priced on access and intuition, is finally large enough and instrumented enough to reward someone who treats it like a pipeline problem. The 2021 vintage suggests that bet is paying. Fund II will say more.