Crossing the Valley

ABOUT BENBen Van Roo is the CEO and Founder of Legion Intelligence. He comes from a military family (his brother is currently on active duty), giving him both personal connection to and deep understanding of the defense community’s needs. His career encompasses a combination of academic rigor and operational execution.ABOUT LEGION INTELLIGENCELegion Intelligence is an applied AI company that connects AI capabilities to the actual workflows, processes, and legacy applications used across the Department of Defense.The Core Problem: AI models have gotten remarkably good, but they remain disconnected from how work actually gets done.The Legion Product: Legion builds the infrastructure that makes any workflow, across any environment, AI-enabled. The company’s offerings runs on edge, cloud, or hybrid environments; connect to legacy applications without requiring expensive system integrator rebuilds; and are model-agnostic (e.g., work with open source, proprietary, and government models). Ben also says they are the first to deploy an agentic workflow platform in IL5 and IL6 classified environments.KEY TAKEAWAYS1. The Plumbing Problem Is the Real ProblemWhen Ben was at Primer in 2017, he watched open source models become “terrifyingly good.” But nobody was actually using them across their workflows. He realized it wasn’t as much about the model as it was about connecting AI to the six or seven systems people actually touch to get work done.2. Find Champions Who See the FutureGeneral Fenton at SOCOM was “light years ahead of everyone, including venture capitalists” in understanding enterprise AI. He wanted to bring AI across the enterprise before ChatGPT made it obvious. Legion secured their initial IDIQ contract because one visionary leader saw where things were going.3. Get Technology Into Users’ Hands ImmediatelyThe most important lesson Ben took from Primer was that feedback cycles are everything. At Legion, the philosophy became: “Any exercise, any event, anything where we can go—we’re going to be there.” Bring your own hardware, do all the work to get there, and prove it works.4. Make Early Technical Bets That Seem ConfusingLegion was first to deploy GenAI on small form factors for national security. First on-prem. First on IL5. First on IL6. These investments seemed odd at the time, when others were just focused on the cloud. But they created defensible positioning as the market evolved.For more on Legion: legionintel.comFollow Ben: https://www.linkedin.com/in/vanroo/Follow Noah: https://www.linkedin.com/in/noahsheinbaum/Subscribe to Crossing the Valley: valleycrossers.com This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.valleycrossers.com

About Ben NicholsonBen is the Chief Business Officer at Ursa Major, bringing over 22 years of government service to the defense tech startup. His career includes 10 years in the Coast Guard, legislative experience as a congressional staffer on appropriations, and corporate leadership at defense giants L3 and Honeywell.At age 50, Nicholson made an unexpected leap from the established defense industrial base to a venture-backed startup, driven by a desire to give back and the realization that his experience and perspective could help bridge the gap between young engineering talent and the realities of defense procurement.A self-described “constitutional geek” who signs half his emails simply “America,” Ben brings an unusual combination of technical credibility, government insight, and entrepreneurial drive to URSA Major’s mission.About Ursa MajorFounded 11 years ago during the space launch “gold rush,” Ursa Major pivoted to focus squarely on defense propulsion, addressing what they see as the long pole in the tent for anything that moves fast in the battlespace.The company is focused on three key mission areas:* Homeland Defense: Hypersonic propulsion systems* Munitions: Solid rocket motors* Space Mobility: Propulsion for orbital applicationsThe Business Model: URSA Major is a products and systems company, not a services shop. They’ve received $250 million of private capital (including a recent $100 million Series E), opened a 400-acre test facility in Northeast Colorado, and can go from clean sheet design to hot fire in 29 days.The Core Innovation: Ursa Major combines additive (3D printing) and agile manufacturing to achieve rapid iteration while building toward scale. Their philosophy is to use printing to lock designs fast and avoid costly mistakes before committing capital to production tooling.Check out the Website: https://ursamajor.com/Follow Ursa Major: https://www.linkedin.com/company/ursamajortech/Follow Ben: https://www.linkedin.com/in/ben-nicholson-666257205Follow Noah: https://www.linkedin.com/in/noahsheinbaum/Subscribe to Crossing the Valley: www.youtube.com/@CrossingTheValley This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.valleycrossers.com

Gecko Robotics — From College Project to Critical InfrastructureABOUT TROY DEMMERTroy is the co-founder and Chief Product Officer of Gecko Robotics. He started the company 12+ years ago as a college senior project at Grove City College in Western Pennsylvania, alongside co-founder Jake Loosararian. Troy came from a medical background and saw an opportunity to bring diagnostic thinking to industrial infrastructure. He personally led Gecko’s expansion into defense starting four years ago, and the company now manufactures robots in Pittsburgh while deploying them across power plants, refineries, naval vessels, and nuclear facilities.ABOUT GECKO ROBOTICSGecko Robotics builds AI-powered climbing robots that inspect and assess the structural health of critical infrastructure. The company started by serving commercial clients — power plants, refineries, and manufacturing facilities — before expanding into defense. Today, their technology inspects ICBM silos, destroyers, submarines (including SSNs and SSBNs), and is moving into new construction for Columbia-class submarines and naval reactors. Sam Altman wrote their first check during their YC batch, and the company turned down an early acquisition offer from an OEM. Gecko manufactures in Pittsburgh, PA and emphasizes keeping robots “employed” rather than sitting unused.TIMESTAMPS[1:10] Origin story: The power plant problem that started it all[2:35] How a college alumni connection led to Gecko’s first customer[4:00] “You can’t build this in a lab” [4:48] Two years in one boiler: Jake’s journey building the first robot[6:33] Forward deployed engineering and the Palantir inspiration[8:02] Finding product-market fit: The YC experience and Sam Altman’s first check[10:00] “Hiding in plain sight” [11:47] Transition to defense: From power plants to national security[13:05] AFWERX and SBIR: Building the foundation for defense work[14:24] What translated from commercial and what required rethinking[16:21] Working with OEMs: Reverse scanning and the first acquisition offer[18:35] First defense asset: The path to inspecting ICBM silos[20:09] Inside the room: Getting robots certified for nuclear environments[21:33] Current state: Scaling across destroyers, submarines, and new construction[25:37] Business model: Robots as capability vs. permanent fixtures[27:00] Manufacturing in Pittsburgh and scaling production[28:07] “95% of robots are unemployed”[29:35] Advice for startups: Think integration from day oneKEY TAKEAWAYS1. Access to the Problem Is the Unfair AdvantageGecko’s origin story embodies the ethos of building with not for a customer. A power plant operator came to their college with a specific problem: 70% asset uptime and constant unplanned downtime. He offered access to his facility. Co-founder Jake spent two years inside a single boiler refining the first robot. Troy is emphatic: “I don’t think you could build a technology like this in a lab. That was the unfair advantage — just access to the problem.”2. Using SBIRs to Build InfrastructureWhen Gecko decided to pursue defense, they started with AFWERX and SBIR programs — Phase 1, and a number of Phase 2 contracts. Beyond the funding, the value was in building what Troy calls “the semblance of a defense company”: registrations, certifications, procurement knowledge, and the operational credibility needed to scale. Jake and Troy recognizing Small Business Innovation Research for what it was: scaffolding for the larger business they wanted to build.3. Target O&M Budgets to Bypass Procurement TimelinesGecko made a strategic decision to enter defense through operations and maintenance cycles rather than new procurement. The reasoning was practical: O&M money already exists and can be spent in-year. “If we had a capability that could be used today to improve an O&M cycle, great. We can buy that in year. We don’t have to set up this longer cycle.” This let them demonstrate value quickly, build relationships, and expand scope from a single destroyer to entire platform classes.4. Start at Forward Bases Where Urgency Is HigherGecko’s first naval deployment wasn’t at a major domestic shipyard — it was at SRF Japan, a forward-deployed repair facility. Troy’s logic: “You’ve got a little less capability at a forward base. The urgency to get things done with creativity and innovation might be a little bit higher than your mother shipyards here back at home.” 5. Design for Integration, Not Just CapabilityTroy’s parting advice cuts to the core of why many defense tech companies struggle to scale: “In order to deploy technologies here, it’s not just pushing code. You’ve got to meet the physical realm where it’s at. If it doesn’t work for the people on the ground, if it doesn’t make their lives easier, you’re gonna get organ rejection.” Gecko succeeded because they thought about tech insertion from day one. FOR MORE:About Gecko: https://www.geckorobotics.com About Troy: https://www.linkedin.com/in/troy-demmer-27037526Crossing the Valley: youtube.com/@crossingthevalley This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.valleycrossers.com

ABOUT DAN SMOOTDan Smoot is the CEO of Vantor, formerly Maxar Intelligence. He comes from a background in high-tech companies and has led major business transformations before. Since taking the helm, Dan has overseen the company’s transition from a publicly traded satellite imaging business to a private, solutions-focused company. He led the controversial decision to retire the Maxar brand — one of the most recognized names in commercial space — in favor of the new Vantor identity. Under his leadership, the company has launched six satellites, dramatically increase recurring revenue, and is expanding aggressively into international markets.ABOUT VANTORVantor (formerly Maxar Intelligence) is one of the world’s leading commercial satellite imagery and geospatial solutions providers. The company operates a constellation of satellites delivering 30-centimeter resolution imagery — the highest commercially available. After being taken private by Advent International and BCI, Vantor sold its manufacturing division and completely transformed its business model from transactional imaging sales to subscription-based solutions. The company’s products include maritime awareness, site monitoring, GPS-denied navigation, and 3D mapping. Vantor claims to be one of only two companies (alongside Google) with global 3D mapping at scale, and serves customers across defense, intelligence, and commercial markets worldwide.KEY TAKEAWAYS1. Sometimes You Have to Kill the Brand to Transform the BusinessMaxar was one of the most recognized names in defense tech, synonymous with exquisite satellite imagery. But that brand equity became a liability. Customers saw “imaging company” and couldn’t see the new capabilities: AI-powered analytics, maritime awareness, GPS-denied navigation, and 3D solutions. The rebrand to Vantor was strategic: “You have to reorient the eyes of the customer to make sure they understand there’s broader modernization happening.” Add the complexity of Maxar Space being sold to Intuitive Machines, and leadership felt keeping the old name would only have created more confusion.2. Going Private Creates Air Cover for TransformationVantor’s transition from transactional sales to 90% recurring revenue didn’t happen under public market pressure. When Advent International took Maxar private, it gave leadership something rare: time and permission to rebuild the business model. Dan is candid about this: “When you go private, you can actually take the time to reformat the business. Getting the sales motion, getting your customers to buy in a different way is not easy.” Today, the company boasts some of that predictability that public markets reward.3. Acquisition Reform Opens Doors for Commercial SolutionsThe recent push for acquisition reform means the government is looking to buy software and modern capabilities differently. Vantor has aligned its business model directly to this shift. Subscription-based solutions fit how the government now wants to buy: modern, updateable, and not locked into elongated contracts.4. Allied “Sovereignty Panic” Is a Massive Growth DriverWith the US stepping back from certain international commitments, allied nations are suddenly realizing they lack organic intelligence capabilities. They’ve been dependent on American systems for decades. Now they’re scrambling. Dan sees this as Vantor’s biggest growth opportunity: “They’ve gotten ‘wow, we don’t really have our own capabilities.’ You can only build that through commercial. It’s almost impossible to have the funding and time to do it bespoke.” This is true across Europe, Asia-Pacific, and beyond.5. Geospatial AI Is a Different Problem Than Language AIDan’s call for startups challenges the current AI hype: “Start thinking about the spatial side, not necessarily the language side. The mathematics is very different. We’ve kind of solved language with things like ChatGPT. Spatial recognition of change on the ground is a whole different way of thinking about data.”TIMESTAMPS[1:19] What brought Dan to RNDF and key takeaways from the day[2:15] Acquisition reform and how it impacts Vantor's strategy[3:24] Why Maxar Intelligence became Vantor [5:42] The role of financial markets[6:28] Breaking down Vantor's core products and capabilities[8:08] Driving automation in geospatial intelligence[8:42] The decision to be more public with capabilities on social media[10:18] GPS-denied navigation: supporting drones in Ukraine[10:43] The China satellite photo showdown[11:43] Non-earth imaging and space domain awareness[12:05] Current phase of business[13:36] International expansion and allied sovereignty needs[14:00] Transitioning to 90% ARR[15:40] Going private [17:31] 3D mapping[19:48] Commercial applications[21:32] Managing petabytes of data [24:03] Dan's call for startupsLearn more: vantor.comFollow Dan: https://www.linkedin.com/in/danielsmoot/For more Crossing the Valley: youtube.com/@CrossingTheValley This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.valleycrossers.com

Nick LaRovere, CEO of Pryzm and former Palantir engineer, discusses transforming government procurement with innovative technology. He shares insights on the company's evolution from basic document processing to dynamic mapping tools that visualize relationships and influences. Nick reflects on early lessons of humility in customer engagement and the significance of the Orion constellations feature for targeted analyses. He also highlights upcoming growth plans, optimism for bipartisan acquisition reform, and the importance of staying mission-focused to drive federal innovation.

Friends,This week, we are spotlighting a new podcast in the defense community, Rebuilding the Fleet.This new pod covers all things Maritime Autonomy and shipbuilding, featuring repeat guest Austin Gray.This episode is a discussion between Austin and HII’s Executive Vice President Eric Chewning, a former investment banker, management consultant at McKinsey, and senior advisor to the Secretary of Defense.All part of our efforts to promote solid content from the next generation of defense disruptors. Have a listen and if you enjoy, we encourage you to follow along and subscribe.Key Topics of Conversation* Workforce Innovation and Outsourcing Strategy: HII’s ambitious plan to scale outsourced work from 2 million to 3 million hours annually, and rebuilding of the sub-tier industrial base across America* Unmanned Systems Leadership: HII’s position as the world’s largest UUV producer, the evolution of the Lionfish program and autonomous launch and recovery capabilities for the Remus vehicle family* Industrial Base and International Partnerships: Analysis of the Korea-Hyundai partnership and how to balance domestic production with international cooperation* Technology Integration Challenges: assessing the state of AI implementation in legacy manufacturing environments and change management challenges in traditional shipbuilding operations* Workforce Development Excellence: HII’s Apprentice School programs and strategies for attracting talent from other industries to maritime* Defense Policy and Budget Outlook: FY26 defense program priorities, evolution of Navy unmanned surface vessel (USV) strategy, and long-term implications for naval force structure and capabilitiesSubscribe to Rebuilding the Fleet at https://www.youtube.com/@RebuildingTheFleet This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.valleycrossers.com

Case Study: Crossing the Valley of Death in Digital ManufacturingAbout Nathan DillerNathan Diller brings a rare combination of operational military experience, government innovation leadership, and private sector execution to the defense technology ecosystem. After 22 years of Air Force service as a test pilot, he transitioned to become Director of AFWERX, where he managed and expanded the SBIR budget, helping hundreds of startups navigate government contracting. His subsequent role on the House Appropriations Defense Subcommittee staff gave him legislative perspective on defense budgeting and acquisition. Now, as Head of Aerospace & Defense at Divergent, he’s applying this comprehensive understanding of the ecosystem to scale manufacturing innovation here in America.About DivergentDivergent has created the Divergent Adaptive Production System (DAPS), an end-to-end software and hardware production system for industrial digital manufacturing. The company’s approach combines digital engineering for structural optimization, software-driven printable design, efficient 3D printing with optimal part stacking, and robotic assembly. Initially proven in the automotive sector (building the world’s fastest production car at 253 mph), Divergent is now applying this technology to aerospace and defense applications. The company has raised approximately $700 million and maintains partnerships with automotive giants like Aston Martin, Bugatti, and McLaren, as well as defense contractors like General Atomics.For more on Divergent: https://www.divergent3d.com/For more Crossing the Valley: valleycrossers.com This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.valleycrossers.com

About EricEric didn’t start as an entrepreneur. After serving 12 years as an infantry officer in the US Army, including deployments during the Iraq and Afghanistan wars, he attended Stanford GSB while still on active duty. Exposure to business school thinking, combined with his frontline understanding of technology pain points, set him on an unexpected path. In 2017, he co-founded Key Square Labs with two technical experts, bringing a military veteran’s perspective to solving a critical but unsexy infrastructure problem.About Key Square LabsKey Square Labs addressed a scaling challenge that few people outside the defense community understood: the ATAK networking problem. Android Team Awareness Kit (ATAK) had become a critical situational awareness app for military users, but it relied on custom hardware with modified firmware to connect to military networks. This worked fine for a small, elite user base, but as adoption grew to hundreds of thousands of users across the joint force, the custom hardware became a bottleneck—expensive, hard to update, and limiting combat effectiveness.Key Square’s innovation was reimagining firmware-level capabilities as application-layer software. This meant any Android device could run ATAK networking without custom hardware modifications. The company operated as a bootstrapped three-person team, funded by R&D contracts and direct sales to eight foreign governments. They never raised venture capital, yet they achieved real-world validation, including deploying with the 75th Ranger Regiment during the Afghanistan withdrawal to solve urgent communication challenges. In 2024, Key Square was acquired by Ditto, a Series B commercial company applying edge computing technology to defense problems.So get out there and enjoy AUSA, CTV readers.We’ll see you next week!For more on Ditto: https://www.ditto.com/Follow Eric: LinkedInFollow Noah: LinkedIn | X This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.valleycrossers.com

David Tuttle, CEO and co-founder of Rune Technologies, brings a wealth of experience from his time as an Army officer and leading JSOC software teams. In this engaging discussion, he shares how his diverse career shaped his mission to modernize military logistics. Tuttle highlights Rune's innovative tactical-edge software, which flips traditional logistics on its head. He discusses the importance of understanding defense budgeting and how his company is tackling challenges for the Army and Marine Corps, aiming to revolutionize the way they operate and make decisions.

About Jeff ColeJeff Cole is the CEO and co-founder of Hidden Level, bringing 20 years of experience developing radar and sensing technology for defense, intelligence, and commercial customers. Before founding Hidden Level, Jeff worked at Saab and SRC (a not-for-profit defense company), where he developed cutting-edge systems for customers including the Army, FAA, and NASA. He also collaborated with commercial giants like Google, Apple, Disney, and Amazon on early drone delivery initiatives, working directly with Astro Teller and Sergei Brin on what would become Wing.Born and raised in Syracuse, New York—an epicenter for radar and electronic warfare technology—Jeff built his expertise in an ecosystem surrounded by companies like Lockheed Martin, Raytheon, and Saab. This background gave him rare insight into both the technical challenges of advanced sensing systems and the procurement realities of government customers. At SRC, he co-founded Griffin Sensors, a wholly-owned subsidiary focused on commercial applications, which shaped his vision for a commercial-first defense company.Jeff’s approach combines technical depth with an entrepreneurial mindset learned from working with fast-moving commercial partners. His guiding principle: “If it wasn’t hard, it wouldn’t be worth doing.”About Hidden LevelHidden Level provides airspace awareness through passive radar and RF sensing technology that detects objects ranging from small, non-emitting drones to fighter jets and balloons. The company’s vertically integrated approach means they design and build everything in-house—software, firmware, phased arrays, and mechanical systems—using a modular “Lego” architecture that enables rapid deployment and reconfiguration.Founded with a commercial-first strategy, Hidden Level initially focused on enabling safe drone delivery and urban air mobility through subscription-based airspace monitoring services. Early customers included NASA, Joby Aviation (formerly Uber Elevate), and commercial enterprises. This commercial foundation proved critical when transitioning to defense applications, as the technology was designed from the start for exportability, interoperability, and rapid scaling.The company’s breakout moment came through a partnership with the U.S. Army. After starting with small SBIR contracts, Hidden Level progressed through an IDIQ vehicle with Booz Allen Hamilton as lead systems integrator, won APFIT funding in May 2023, and achieved program of record status under urgent capabilities in January 2024—just 18 months from initial prototype to fielded production systems. During the December 2024 drone crisis in New York, Hidden Level deployed sensors at Stewart International Guard Base in under 24 hours, enabling the apprehension of unauthorized drone operators within minutes.The company has raised over $100 million from investors including Quest Ventures, DFJ, Costanoa Ventyres, Washington Harbor, Lockheed Martin Ventures, and Booz Allen Ventures. With 130+ employees and growing, Hidden Level is scaling both commercial infrastructure deployments across U.S. cities and defense applications globally.Key Takeaways1. Commercial-first beats defense-to-commercial for dual-use companiesHidden Level’s approach didn’t take defense technology and try to commercialize it; instead, they built with commercial intent from day one and then adapt to defense needs. Hidden Level designed for subscription models, exportability, and interoperability—requirements that made government adoption easier, not harder. The modular architecture that enables rapid deployment in commercial settings (like the 24-hour Stewart AFB installation) directly translated to defense value. This approach avoids the vendor lock-in and compliance baggage that makes defense-to-commercial transitions so difficult.2. Reputation and relationships create momentum that capital alone cannotBefore Hidden Level existed, Jeff and his team had delivered advanced radar systems to demanding customers for two decades. This track record meant NASA, Joby, and Army customers believed in their ability to execute even when working from a basement. When first investor Tom Moss tripled his commitment within 48 hours and introduced Jeff to other VCs, it wasn’t just about the technology—it was about backing a team with proven delivery capability. For defense tech founders, past performance and domain expertise can be more valuable than a perfect pitch deck.3. APFIT and other bridge funds are really important to bridge the valley of deathHidden Level used small SBIR awards to maintain customer relationships, moved to an IDIQ through partnership with Booz Allen Hamilton, then leveraged APFIT funding to procure systems when the Army customer had validated demand but lacked budget. This represents a careful understanding of which funding mechanisms match which stage of technical maturity and customer pull. APFIT worked because Hidden Level had already proven the technology and had an evangelizing customer; it wouldn’t have worked two years earlier.4. Partnerships with primes require clear-eyed understanding of incentives and termsJeff’s advice on working with integrators like Booz Allen Hamilton and strategic investors like Lockheed Martin is notably nuanced. These relationships can be powerful but require understanding contractual vehicles (FAR parts 12 vs. 15), IP ownership, colors of money, and compliance requirements upfront. The relationship with Booz Allen worked because roles were clear—they were the lead systems integrator doing C2, Hidden Level provided sensors and integrated into their architecture. Going in eyes-wide-open about what the partnership actually entails prevents later frustration about doors not opening or unexpected IP constraints.5. Intentional product architecture enables speed at scaleHidden Level’s “Lego modular” design philosophy is a strategic choice that enabled their 18-month prototype-to-production timeline and 24-hour deployment capability. Components designed for one product line work across different applications, reducing development time for new variants when the Army wanted a vehicle-mounted system half the original size. This modularity also supports the dual-use model: the same core technology serves commercial airspace monitoring subscriptions and military counter-drone applications. Speed in defense tech isn’t about cutting corners—it’s about making architectural choices that create optionality and reduce integration friction from the start.For more on Hidden Level: Website | LinkedInFor more Crossing the Valley: Substack | YouTube | LinkedIn Follow Jeff: LinkedIn Follow Noah: LinkedIn | X This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.valleycrossers.com