If being successful requires 1000’s of decisions/actions to go your way, it wasn’t luck

My @MarathonMP partner @ChaseAPackard, with his prior investments in Revolut and Mercury, among others, knows more about what best in class looks like for neobamks , than most anyone else in the world.

"Neobanks are easy to run"

Leading Wafer's Series A tl;dr @MarathonMP is co-leading Wafer’s $40M Series A with our friends at @Chemistry, with participation from @AMD Ventures, @Wing_VC, @outsetcap, @fiftyyears, @ycombinator, and several amazing angels @JeffDean (CEO, @DiscoLoopAI), @rauchg (CEO @vercel), @andyfang (CTO, @DoorDash), @akothari (COO, @NotionHQ), @kvogt (CEO, Bot), @eastdakota (CEO, @Cloudflare) and deepgramscott (CEO, @DeepgramAI), among others. Everything will be inference. Inference is and will be the single largest market within AI. It's the white hot center within the white hot center of AI. Today, to make inference work well, highly specialized engineers manually tune models, serving engines, kernels, and hardware. The work is slow, expensive, and often repeated when traffic changes, models improve, or new chips arrive. Wafer has built a fundamentally difference inference architecture from the ground up. Wafer's agents continuously study production traffic, identify bottlenecks, test changes, verify the results, and deploy the best-performing configurations. Simply put, Wafer builds AI that improves AI. And boy, does it improve AI. Customer after customer we spoke with, raved about Wafer's performance compared to incumbent inference providers. A very large platform said that they plan to move 100% of their inference volume to Wafer as soon as possible. Wafer cofounders @gpuemi and @gpusteve have a deep understanding of this problem through working on GPU and kernel optimization for an extended period of time. One of the most impressive things about Wafer - besides their product - is the exceptional team that Emilio and Steven have built. They have built a distinct, compelling, talent-dense culture and created an incredibly high bar for hiring that i've observed in every generational company I've been part of. Wafer has already doubled in size in the weeks since we invested, and continues growing at a rate that puts them in rarefied territory. More importantly, their product is unique, differentiated and we continue to get unsolicited raves from their customers. We’re incredibly excited to support the @wafer_ai team as they pursue their mission to maximize intelligence per watt and build a generational company that will matter for decades. PS: h/t my partner @ChaseAPackard for the meme below.

Is AMD catching up to Nvidia performance? @wafer_ai CEO Emilio Andere explains his experience: "AMD, if you optimize the software correctly, can actually run at similar performance than Nvidia.” Read more: thein.fo/4gwuvLF

Wafer (@wafer_ai) is building a fast AI inference cloud that uses agents to optimize GPUs and run open-source models at industry-leading speeds. The result is the low cost of open source with the latency of a much smaller model. Four months after launching the cloud, they went from zero to $8 million in ARR and just raised a $40M Series A. In this episode of Founder Firesides, YC's @snowmaker sits down with Wafer co-founders @gpusteve and @gpuemi to talk about going from UChicago undergrads to building one of the fastest-growing AI infrastructure companies in just over a year. They share how a college side project became Wafer, how they got GLM 5.2 running 2-3x faster than anyone else in the market, and why speed, not just cost, is pushing enterprises toward open-source models. 00:48 — What Wafer Does 01:35 — Going From $0 to $8M ARR in Four Months 04:21 — Scaling Faster Than They Could Keep Up 09:14 — How AI Makes AI Faster 11:21 — The Pivot That Changed Everything 14:41 — Why Open-Source Models Are Taking Off 18:03 — From a College Side Project to Wafer 23:33 — Advice for Starting and Joining Startups

We at @MarathonMP are proud to lead @wafer_ai's Series A alongside @chemistry, @ycombinator, @Wing_VC, @fiftyyears, @AMD, and a great group of angels. Wafer has built fundamentally new inference architecture from the ground up. Wafer's agents continuously learn from production traffic, identify bottlenecks, test changes, verify the results, and deploy the best-performing configurations. In short, they use AI to drastically improve AI. I met @gpusteve four years ago on campus - the friend who introduced us called him the smartest person he'd ever met, and after four years I still haven't found evidence to the contrary. Same goes for his co-founder @gpuemi. Backing @UChicago founders has been a long-time goal of mine; very glad it's these two. CC @gokulr @AlexGorgoni @MBGilroy @GraceGEverett

Wafer, a year-old inference provider that uses Nvidia and non-Nvidia chips, turned down acquisition offers to raise $40m in new funding at a $200m+ valuation. Marathon and Chemistry co-led. For more, check out this morning's AI Agenda: theinformation.com/newslette…

INVESTOR VALUE ADD My partner @ChaseAPackard and I were recently discussing investor value add. Chase made the point that for early stage (pre-Series B) companies, investors who add zero value might as well add negative value. I think he’s right. Before Series B, a company has very little slack. The founders are still assembling the team, finding repeatable growth, raising capital, and making decisions that can materially change the outcome. Every investor occupies scarce space: on the cap table, in the founder’s calendar, and sometimes in the boardroom. A passive investor still creates work. They need updates. They may ask for calls at the wrong moments. Their advice has to be evaluated, even when it comes without the operating context to make it useful. There’s also an opportunity cost. Every allocation given to a passive investor could have gone to someone who helps the company: • Recruit a key employee • Win a critical customer • Prepare for the next financing • Work through a decision when the answer is unclear The standard shouldn’t be whether an investor is pleasant, responsive, or well known. The useful test is whether the company would miss them if they disappeared from the cap table tomorrow. Value add also changes by stage. At the earliest stages, founders often need direct help: recruiting, customer introductions, fundraising. As the company grows, judgment becomes more important. The highest-impact investor may be the person who helps the CEO see one consequential decision more clearly. Founders should evaluate investors with the same rigor they apply to early stage employees or executive hires. 1. CONCRETE EXAMPLES: Ask them for specific examples of how they helped companies at your stage. 2. BACKCHANNEL: Speak with founders who went through hard periods with them (and also to validate #1 above). Tough periods bring out the real character of people. 3. WORK PRODUCT: Establish if they can contribute, BEFORE you sign the termsheet. Can they actually intro you to a customer or an employee, or help with a key decision? Early stage founders: a passive investor carries a real cost. Choose people who will earn their place on the cap table.

Two years ago, I found myself seated next to a prominent VC at dinner. We got into a debate about the future of the Office of the CFO. That particular VC was deeply pessimistic on the category. Their view was that these products, and ERPs in particular, were simply too entrenched to disrupt. I strongly disagreed. I made the case for @RilletHQ: companies willing to rebuild core financial infrastructure instead of accepting the limitations of systems designed decades ago. The naysayers often confuse “difficult” with “impossible.” Yes, replacing an ERP is difficult. So is changing how companies run payroll, move money, close their books, and manage their workforces. But difficulty does not protect an incumbent forever. It only raises the bar for the builders who come after it. @Rillet announced a $100 million Series C led by @ICONIQCapital at a $1 billion valuation. Congratulations to Nicolas Kopp and the entire Rillet team. We’re proud to have Rillet as a Niural customer and delighted to support a company rebuilding one of the most important systems inside the Office of the CFO. And, on a personal note, I’m happy to have been on the right side of that dinner conversation. Change happens, no matter how difficult it looks before it does.

@niuralHQ is launching its SF office, and bringing NFL Hall of Famer @LarryFitzgerald to the party. Come join us!

Only @niuralHQ is AI native payroll. Everyone else is sparkling payroll wrappers.

shoutout to @ChaseAPackard for breaking the news on TITV today re: airtable's spinoff before its recent sale! @validapau @julia_hornstein and I have even more exclusive details in our dealmaker column on the deal: theinformation.com/newslette…

Airtable’s $1.3B sale may be just the “opening act” in AI-driven SaaS consolidation. "The February 2026 SaaS apocalypse wiped out almost $300 billion of market cap in a week." "Because AI worked, customers are cutting seats as AI as more agents take more of this work." — @MarathonMP Founding Partner @ChaseAPackard

Today on TITV: -Bending Spoons to acquire Airtable for $1.3B | @ChaseAPackard, Founding Partner at Marathon; @aaref, Partner at Bain Capital Ventures -Why a SpaceX-Tesla merger could benefit shareholders | @AnitaRamaswamy -Coinbase, others seek Claude Code alternative | @LauraBratton5 -Goodfire launches AI research platform ‘Silico’ to public | @eric_ho, Co-Founder and CEO of Goodfire 📺 Tune in at 10 am PT / 1 pm ET on thein.fo/4fCtW2j

🚨 BREAKING Kimi K3 Fast⚡️is live on @OpenRouter the fastest serverless Kimi K3 in the world pick @wafer_ai as your provider: openrouter.ai/moonshotai/kim…