Defense Startups, Crossing the Valley, and Steam Studios with Noah Sheinbaum

Startups can’t afford to be wrong on an 18-month cycle—and neither can the services or the acquisition system that supports them. Callye and Noah deconstruct the so-called “valley of death,” separating structural reality from self-inflicted pain and inexperience. They dig into how narrative, incentives, and tighter demo cycles can compress risk and get real capability into the hands of operators faster.

Topics

• Why the “valley of death” is often more mirage than destiny—and where it’s brutally real
• How information, narrative, and media shape behavior in defense and critical industries
• Kform’s evolution from a third-generation machine shop to a commercialization partner for defense startups
• Competition-based drone events and what they reveal about the future of acquisition
• Using rapid demos and shared customers to shrink timelines from years to weeks

Takeaways

• The valley of death isn’t an inescapable trap; it’s a series of predictable transitions that punish inexperience, poor capital efficiency, and lack of customer focus.
• In defense, you may only get one meaningful shot every 12–18 months, so choices around SBIRs, sponsors, and end users are not “free”—they are existential.
• Tight feedback loops—design, build, demo, adjust—run on weeks instead of years can derisk programs, align incentives, and help both startups and the government “be slightly wrong and get less wrong” much faster.

Timestamped Highlights

[00:00] - Why startups (and DoD) can’t afford 18-month mistakes
[02:34] - Rethinking the “valley of death” and reclaiming agency
[05:24] - A contrarian view: the valley as mirage and self-inflicted pain
[09:02] - “SBIR is not free money” and the cost of choosing the wrong sponsor
[11:54] - Defense tech jobs, new attention, and why this moment matters
[17:33] - Kform’s roots: from Navy machinist and Circle K to Kform
[23:55] - Building Kform’s playbook: the “team behind your team”
[26:29] - Steam Studios and competition-based drones as a new acquisition pattern
[31:20] - Tight demo loops: slightly wrong every two weeks vs. catastrophically wrong in 18 months
[36:57] - Policy shifts, acquisition reform, and closing reflections

Resources & Links

• Kform — https://kform.com
• Crossing the Valley (podcast) — https://www.frontdoordefense.com/podcast
• Front Door Defense Jobs — https://www.frontdoordefense.com
• Steam Studios — https://www.steamstudio.io
• U.S. National Drone Association — https://www.usnda.org

“We can’t afford to be wrong in 18 months. We can afford to be slightly wrong—and getting less wrong—every two weeks.”

To partner with Kform, explore collaboration, or bring a product from prototype to field faster, visit kform.com and reach out to the team. If this episode resonates, share it with a founder, operator, or investor who cares about actually fielding capability—not just talking about it.