Pioneering a better space launch system: why we invested in Moonshot
Gil Dibner

Origin story. In August 2023, an incredible entrepreneur named Shahar Bahiri emailed me a slide deck for a company he was building called Moonshot. The idea sounded completely crazy at the time: an electromagnetic space launch system designed to deliver payloads to orbit at radically reduced cost—effectively, a maglev train to space. Listed as co-founder was Hilla Haddad Chelmnik. Although she was only part-time at Moonshot then, her presence on the team spoke volumes about Shahar’s entrepreneurial capabilities and the fundamental appeal of his vision. Hilla is a powerhouse aerospace engineer who previously held senior civilian government roles in Israel, where she oversaw major components of the development and deployment of the Iron Dome and Arrow missile systems.

My first response was to send the deck to Fred Simon, a Venture Partner at Angular Ventures and the former Chief Architect of JFrog, a $6B publicly traded company he founded in 2008. Having worked closely with Fred for several years, I knew he was going deep—incredibly deep—on spacetech. A few days and several questions later, Fred and I met Shahar. A few months later, we invested in Moonshot, leading the company’s $2.5M pre-seed round alongside Fred himself.

Fred, for his part, had scrutinized the math and physics. He reviewed the energy density requirements and material science constraints. He examined the feasibility of releasing massive amounts of energy with the precise timing required to propel a vehicle through a coilgun at hypersonic speeds. Based on the fundamental physics, he became convinced that if this worked, the impact on space launch economics would be dramatic. Most crucially, Fred—like David and me at Angular—became convinced that the team Shahar and Hilla were assembling could actually pull this off. Even more significantly, Fred decided to join the company full-time. Having already built a $6B devtools SaaS company, Fred felt his next calling was to build something even bigger.

The risk-reward trade-off in deep tech. Angular does a lot of deep tech investing—an area we often call “weird tech.” The more radical a technology becomes, the larger the potential outcome must be to justify the engineering risk and basic science hurdles. In Moonshot’s case, that equation balances perfectly. The engineering risks here are massive; no one has ever launched anything into orbit using this method. But if Moonshot succeeds, the upside is enormous. They are poised to fundamentally rewrite the economics of space launch. For a reward of that scale, we are willing to accept uncommonly high risk.

Rewriting the rocket equation. The Rocket Equation is a fundamental limitation in space travel governing the efficiency of propulsion systems that rely on chemical fuel. It dictates that for a conventional rocket to reach orbital velocity, the vast majority of its mass must be chemical propellant. Essentially, you need fuel to lift the payload, but you also need fuel to lift the fuel itself. Consequently, modern chemical rockets have a notoriously low payload capacity, typically limiting the actual cargo to only about 4% of the vehicle's total mass. This results in inefficient and costly logistics; for instance, launching a kilogram of water costs the same as launching a kilogram of high-tech equipment.

Moonshot sidesteps the traditional rocket equation by employing a kinetic launch system based on electromagnetic acceleration. (A small rocket is used only for the final stage.) The technology uses a contactless electromagnetic accelerator as the first stage, propelling payloads up to 8 kilometers per second using electricity to provide the majority of the energy required for orbit. By eliminating the heavy fuel tanks traditionally needed for liftoff, Moonshot drastically reduces the spacecraft's mass. This revolutionary approach boosts the payload fraction from the standard 4% to over 45%—potentially targeting 60%—effectively achieving a theoretical 10X reduction in spacecraft mass and a 10X increase in payload capacity. Eliminating chemical fuel as the primary energy source enables Moonshot to overcome the limitations of the classical rocket equation and offer a highly affordable supply chain delivery system. Needless to say, this is a complete game-changer for the space economy.

A unique opportunity. Moonshot recently announced their $12M seed funding round. We were delighted to participate, and we are thrilled to have backed the company from the pre-seed stage. We are grateful to Shahar for his ongoing leadership and courage in moving this process forward. We are thrilled to back Fred again, this time in his new incarnation as a spacetech pioneer. And we are beyond excited that Hilla has joined the company full-time as CEO.

Together, these three founders have assembled a truly extraordinary team. They have recently been joined by Caryn Schenewerk (formerly of SpaceX and Relativity Space) as U.S. GM, and Alon Ushpiz (former Director General of Israel’s Ministry of Foreign Affairs) as Head of International Strategy. The engineering team is of equally high calibre. Moonshot has a real opportunity to revolutionize space launch forever.

The path to commercialization. When we first invested, the company believed they might generate interim revenue by finding near-term applications for a smaller version of their acceleration system. We viewed this as an interesting potential upside that was unlikely to materialize. The company is now well on its way to proving us wrong, providing a crucial national-scale hypersonic test capability. If things go to plan, a working initial version of the Moonshot accelerator will be generating revenue by 2027.

A distinctly Israeli opportunity. There is something distinctly Israeli about Moonshot worth noting. Israel has always faced disproportionate challenges unprecedented in scale and complexity. When Israel first proposed building an anti-missile system in 2007, the idea was widely considered impossible. The Americans reluctantly agreed to fund part of the project but did not expect it to work. Yet, over the past two years of conflict, Israel fielded the world's most advanced multi-layered defense system, intercepting thousands of projectiles—from short-range rockets to ballistic missiles in space—with a success rate of over 90%, saving countless civilian lives. Today, governments around the world are scrambling to catch up, and many, such as Germany, are buying these systems directly from Israeli producers. There is an optimistic outcome from Israel's reality of constant asymmetric threat: Israeli entrepreneurs know how to deliver unprecedented systems that simply must work. There is no alternative. This has trained a generation of entrepreneurs not just to believe in the impossible, but to deliver it ahead of schedule and under budget. Moonshot is a direct continuation of that legacy. We couldn’t be more excited for what's next!

Hard tech. Hard markets. We see Moonshot as a great example of what radically contrarian venture capital must be. If you are building something crazy, weird, or impossible - please let us know!

The AI-Native Enterprise Playbook
Ten real-time observations on a rapidly evolving playing field

No More Painting by Numbers
It’s the end of the “SaaS playbook.

The Age of Artisanal Software May Finally be Over
Every wave of technological innovation has been catalyzed by the cost of something expensive trending to zero. Now that’s happening to software.

Founders as Experiment Designers
David on why founders should run everything as an experiment.

ENTERPRISE/TECH NEWS

Long live CRM. Jamin Ball, over at Clouded Judgment, pushes back on the “agents replace systems of record” meme by redefining the core job as where truth lives (i.e. the canonical answer an automated workflow should trust when definitions conflict). As agents become cross-system and action-oriented, the failure mode shifts from “model is wrong” to “agent pulled the wrong value from the wrong place,” which makes explicit semantic contracts, precedence rules, and governance more important. In other words, the durable wedge won’t be a chat UI sitting atop someone else’s data, but the “truth control plane” layer (warehouse/lakehouse + semantic layer + policy) that agents rely on to read/write state safely. That’s where stickiness (and multiples) will concentrate.

OpenAI pivoting to the enterprise. At a private lunch with editors this week, Sam Altman said OpenAI will make selling to enterprises its top priority in 2026. The move reflects a growing awareness that application / tech diffusion, not model training, is now the bottleneck, and that enterprise AI (forecast at $37.5B next year, up from near-zero in 2022) is the surest way to finance OpenAI’s $1.4T infrastructure ambitions. Also an indication that Anthropic’s focus on enterprise from day one seems to be paying dividends and OpenAI wants a bit of that pie.

HOW TO STARTUP

Software’s next act. Kyle Harrison, over at Contrary Research, wrote an interesting deep dive on “composable software,” published last week. He argues that AI won’t kill SaaS so much as dissolve the monolith. Software will shift from fixed apps + plugins to interoperable primitives that can be assembled (and eventually assembled by agents) into bespoke workflows. He frames the last 70 years as a cost-vs-customization tradeoff, and claims that AI finally makes “diagonal” software viable (i.e. horizontal building blocks that stretch into long-tail verticals without giant codebases).

The human bottleneck. On Lenny’s podcast this past week, Alexander Embiricos, OpenAI Codex product lead, frames Codex as the first real “software engineering teammate.” He claims Codex has grown ~20x since GPT-5, is now the most-served coding model, and is already accelerating OpenAI internally (e.g., a Sora Android app shipped from zero to public in 28 days, and Codex increasingly “babysits” its own training runs). As Embiricos says, increasingly, humans on the team are the biggest bottleneck (not the model itself).

Design to code. Cursor’s launch has to sting for the many, many design to code startups out there, as well as Lovable/Bolt/Base44. This is a real category, but will be interesting to see if it ends up being eaten by the IDE.

HOW TO VENTURE

AI growth is out of this world. That’s one takeaway from Meritech’s 2025 market update. Give it a watch here. Or from ElevenLab’s recent $14M ARR day. 

What if ARR isn’t sticky? Kyle Poyar, over at Growth Unhinged, argues that AI startups are breaking the mental model of ARR as durable and sticky. Lots of AI “ARR” is really experimental spend, opt-out contracts, usage/success fees, or services-heavy revenue that behaves nothing like classic SaaS. In a dataset of about 3,500 companies, AI-native retention looks starkly worse than B2B SaaS (median 40% GRR / 48% NRR vs. much higher SaaS stickiness), with a huge bifurcation by price point (namely, $250/mo AI products look SaaS-like, while sub-$50 “wrapper” products churn fast). I wonder if, for their next financing pounds, companies will need to start proving “durable ARR” via workflow depth, upmarket contracts, and tighter adoption loops (often with a margin trade via forward-deployed services).

The power of being a generalist. Aydin from Felicis taking us all to school on Mario Gabriele’s podcast last week. The whole interview is worth a listen, but wanted to highlight to items. First, his take on pricing: you can be “10x wrong” on entry price if you were even more wrong (conservatively) on future revenue growth (30–100x+). (The mistake most investors make is over-weighting price and under-weighting the variance in growth.) Second, he argues generalists aren’t spraying/praying, but starting with a bigger universe of potential investments. As he says: specialists start with a constrained funnel and end up buying #7 in-category because the #1s are gone. Generalists can pick the absolute best deals (across categories) instead of the locally best deals (within a sector). Just two of many insights. Give the whole conversation a listen!

Medida raises $4M seed to turn smartphone scans into millimeter-accurate renovation measurements.

FalkorDB hits the number 1 trending spot for C projects on GitHub.

Aquant
Senior Enterprise Account Executive (NYC, Hybrid)

FalkorDB
Core Developer (Tel Aviv, Hybrid)

Motorica
Account Executive (Remote)