Sponsored by

Good morning, good morning, good morning! Hope everyone's caffeinated and quarterly-forecasted. 🕶️

 

Here's the thing about this week: someone put a live GPU in orbit, and suddenly every hard constraint just became interesting again. Rockets are booking out. Launch windows matter. Engineers who can survive radiation just doubled their rate cards.

Meanwhile, back on Earth, the biggest infrastructure bets are being sized in tens of billions because frontier AI stopped fitting inside normal buildings. The pattern is clear,scarcity is strategic now, and the founders booking capacity early are playing a different game than the ones still optimizing pitch decks.

The founder’s dashboard / Your quick roadmap

FOUNDER BOARD



  • 💡 Raj Toleti’s 1-rule took Andor Health to a $100M valuation and profit, and sold 4 startups by building only what earns.

  • 📊 A $1B‑backed founder on what VCs actually want—and why clean finances beat last‑minute rounds—not panic cash‑ins.

  • 🗂️ Ditch the daily to-do—task batching trims context-switching and lifts deep-work hours and output.

  • 💸 Martin Casado sees VC tilt to tiny teams deploying big funds as startups scale on lean ops, not 100+ hires.

  • 🛠️ Use seed cash to erase the 1 deal-killing objection investors fear most, and tighten your Series A case.

  • 🤖 AI startups Greyparrot, Emesent, SoundHealth, Forge Industries and Cascade raise $63.6M to bring physical AI to recycling, mining, breathing and construction bids.

Startup Investor Finder

Filter by stage, industry, and location, organize outreach, and manage your fundraising pipeline in one streamlined workspace. No random spreadsheets, no scattered LinkedIn tabs; just a focused system to turn research into real fundraising traction.

RUSHIN' ROULETTE



Four bullets of updates

  1. 📈 AI data startup Micro1 hits $500M annualized gross run rate selling training datasets to AI model developers. The milestone shows data vendors, not just chipmakers and model labs, are capturing spend as enterprises race to build advanced models, making the AI training data boom a key infrastructure layer and possible bottleneck.

  2. 🔓 Apollo Global Management, private equity giant managing capital for pension and sovereign funds, confirms a data breach amid a hacking wave targeting financial firms. The incident shows that even top-tier finance players are getting hit shortly after Google's hacking warning, raising the bar for portfolio companies’ security, vendor diligence, and incident playbooks.

  3. 🚨 Dutch data protection authority fines ride-hailing giant Uber $966 million for using automated systems to suspend drivers. The case treats opaque algorithmic HR tools as a major compliance risk, signaling that large-scale automated worker suspensions without human review can trigger billion-dollar penalties and stricter oversight of AI in employment decisions.

  4. 💸 Broadcom, the data-center chipmaker, is lining up more than $60 billion in debt to build AI processors for Anthropic and other labs, with facilities sized for over 20 gigawatts of power by 2028. The move shows how frontier AI now demands up to $100 billion projects that resemble utility-scale infrastructure, not traditional chip orders.

SPONSOR



Stop making AI decisions in the dark.

Leadership is asking: are we getting value from AI? Which tools are worth the spend? Where are we exposed? Right now, most teams have no idea.

You get a complete picture of how your organization uses AI, automatically categorized into custom tasks and use cases.

You’ll see the projects being worked on, who’s using what tools, where AI investments are driving value, and where employees are engaging in risky behavior.

CIOs can rationalize spending and cut wasted licenses. CISOs can pinpoint where risk exists and neutralize it. AI committees can show exactly how their efforts are paying off.

STARTUP NEWS



Starcloud raises $250M to bet on AI, space, and scarce rockets

Starcloud just did the most audacious thing in tech: it decided Earth was too small. In case you missed it, this startup is stuffing Nvidia GPUs into orbit and running AI inference in the vacuum of space. Not a pitch deck. A live H100 is already up there, the only one on the planet, or rather, off it.

So how big is the money?

Big enough to make a terrestrial data center blush. Starcloud quietly closed a $250 million Series A extension led by Manhattan West Ventures, with Nvidia writing a check for $25 million. Total Series A now sits at $420 million, valuation $2.3 billion. Five months ago they raised $170 million. That is not growth. That is escape velocity.

Meanwhile, on the logistics layer… the story gets grim. Rockets are running out. SpaceX's trusty Falcon 9 sunsets in 2028, and the understudies (Blue Origin's New Glenn, ULA's Vulcan, Rocket Lab's Neutron) aren't flying regularly. CEO Philip Johnston isn't buying capital. He's buying launch slots before the door slams shut.

The less obvious ripple… hits your hiring plan. Every radiation-hardened hardware engineer just got very, very expensive.

What should founders actually do?

Lock the long-pole resources now. Launch windows and manufacturing slots are the new oil. Pre-book, or watch late 2026 heavy-lift schedules decide your fate for you.

STARTUP TV



How to build a Company YouTube Channel (and not die trying)

Most companies approach YouTube like another marketing channel. That's usually the first mistake.

Running a YouTube channel for a tech company is very different from running a creator channel. You need the right team, the right host, the right type of content—and you need to know what you're actually trying to accomplish before spending thousands of dollars producing videos.

In this episode, Caya breaks down what we've learned from building Slidebean's channel over the last eight years: why the Producer + Host relationship matters, when a founder should be the face of the channel, how to structure a YouTube team, and why subscribers are mostly a vanity metric.

FOUNDER BRIEF


What founders clicked on most in recent issues

🚕 McDonald’s pilots Escalate app at 7 Maryland locations, slashing shift callout costs by $35,000 per month via Uber vouchers.

Under the pilot, frontline staff file last-minute requests through Escalate, which verifies shift details and instantly issues single-use Uber ride credits or prepaid cards for childcare and groceries. Tying eligibility to upcoming schedules kept abuse low while materially shrinking no-shows inside one franchisee’s P&L.

🚀 Kog turbocharges Nvidia H200 and AMD MI300X GPUs to 3,000 TPS—aiming for 30x faster LLM inference via software.

Under the hood, Kog fuses the entire decode step into a single monokernel that runs across 8 GPUs and delays tensor-parallel all-reduces by two layers, keeping communication almost hidden behind weight streaming. That same layout is now being adapted from the 2B Laneformer preview toward 10x-accelerated frontier-scale LLMs, with a first large-model milestone targeted for September.

🤖 Startup founders report 40% increase in AI-generated pitch decks, fueling investor backlash over generic presentations.

Investors now flag decks that lean on templated “Why now,” hockey-stick forecasts, and buzzword-packed market slides that read like the same prompt was reused, and several funds have quietly added “obvious AI authorship” as a negative signal in their internal scoring rubrics.

🚪 Altour rejects 22% of inbound contracts in Q1 2024 to defend core quality and prevent capacity overrun.

Leadership now runs every incoming engagement through a scoring model that flags red-line risks like rushed onboarding, misaligned expectations, or overextended delivery teams. Deals that don’t clear the bar are declined even when they come from marquee brands or would materially boost quarterly revenue.

STARTUP EVENTS



Startup Events and Deadlines

  1. Antler Continental Europe: Fall 2026 | Sep 14 | Apply

  2. Startup Summit Lisbon 2026 | Sep 17 | Attend

  3. Slush 2026 | Nov 18-19 | Attend

How did we do?

Your feedback fuels us.

Login or Subscribe to participate