NODA AI co-founder and CEO Philong Duong says the US Department of War awarded his two-year-old defense software business a $100 million contract to coordinate autonomous forces across military services and hardware vendors. A related federal notice describes something different: a prize challenge offering up to $100 million across multiple awards, rather than a single contract for NODA.
The Austin-based company made its claim on August 17 through PR Newswire. NODA described the purported award as a vehicle for scaling its mission-command software across the Joint Force. The release did not identify the awarding office, procurement vehicle, performance period, delivery milestones or amount initially obligated.
The Defense Innovation Unit's January announcement says DIU, the Defense Autonomous Warfare Group and the US Navy opened an Autonomous Vehicle Orchestrator prize challenge offering up to $100 million across multiple awards. The program seeks prototypes capable of coordinating autonomous systems from different manufacturers at fleet level. It does not identify a $100 million contract for NODA. The Small Business Administration's official portfolio lists a related $74,504 Phase I award to NODA for vendor-agnostic orchestration of heterogeneous autonomous systems.
Duong arrived at the problem through two careers inside military aviation and enterprise software. He spent over a decade on active duty as a Marine Corps helicopter and tiltrotor pilot and flight-line maintenance officer before joining C3 AI. In a 2020 account of his work, Duong said C3 AI's defense maintenance products covered more than 1,200 aircraft across seven platforms. He holds a computer science degree from Elon University and an MBA from Columbia University.
Duong later led C3 AI's defense and intelligence products operation with fellow Marine veteran Dave Corbett. The pair founded NODA AI in 2024 around a problem they had seen from both sides: the military was buying increasingly capable autonomous vehicles while the software for commanding mixed fleets remained divided by manufacturer, platform and mission.
That founder history explains NODA's product strategy. Duong and Corbett are not trying to build another drone, boat or ground vehicle. They want NODA to become the control layer that tells hardware from different suppliers how to work toward a commander's desired outcome.
The software layer above the drones
NODA's core product, URZA, interprets a commander's intent and matches it with available kinetic, non-kinetic, intelligence, surveillance and reconnaissance assets. A second component, which NODA calls its software-defined-effects engine, determines how combinations of manned and unmanned systems could produce the requested effect.
Duong framed basic drone tasking as an established capability. NODA's harder technical bet is software that can adjust assignments as vehicles fail, conditions change or communications deteriorate. NODA says missions can continue under degraded or denied network conditions and that its integration method requires no hardware or software changes to government or manufacturer platforms. Those operational claims remain NODA's and require government testing or deployment data for independent validation.
NODA says URZA has been demonstrated across 10 platforms and connects through MELD, its open integration standard, to more than 30 government and original-equipment-manufacturer systems. NODA lists integrations and algorithms spanning communications management, sensor classification, electronic attack, urban surveillance and the handoff between sensors and effectors.
Any federal deployment would turn that architecture into an adoption challenge. NODA would have to integrate systems built under different acquisition programs, with different interfaces, security rules and levels of autonomy. Its cross-vendor pitch becomes more valuable as the number of connected platforms grows, while every additional integration creates another dependency that must work under battlefield conditions.
The federal prize challenge closely matches NODA's pitch. DIU is seeking software that can translate a commander's voice, text or physical input into machine execution while remaining independent of any one vehicle supplier. The program anticipates multiple awards. The Los Angeles Times reported that NODA was providing orchestration software for an Applied Intuition-Sierra Nevada Corporation team in the competition, but neither DIU's notice nor the public award record establishes that NODA received the full $100 million.
From an unpaid prototype to a federal competition
NODA's path began with a small group following Duong out of C3 AI. In a June 2026 podcast profile, Duong described senior engineers, data scientists and product designers leaving with him before NODA had settled on a product thesis. The profile said much of the early group worked for close to a year without pay after expected financing fell through.
The group built and demonstrated URZA within months, according to Bessemer Venture Partners' account. Bessemer led NODA's $25 million Series A on February 26, with Booz Allen Ventures, Draper Associates, Bloomberg Beta, Alumni Ventures, Outlander VC and Crosslink Capital participating. NODA did not disclose a valuation.
That financing was meant to fund government milestones and expand NODA's integrations. Roughly six months later, NODA's press release presented the federal challenge's $100 million ceiling as a contract awarded to the company, although DIU describes a multi-award competition.
NODA's news page also points to a $10 million Pentagon award announced on July 23 for work involving LUCAS, an attritable unmanned strike system, and other platforms. NODA has not specified how that work relates to its participation in the orchestrator challenge. Defense contract headlines can describe maximum program values, task-order ceilings or funded awards with very different revenue implications.
NODA's commercial position rests on working across suppliers rather than owning the underlying machines. Hardware manufacturers have incentives to make their own vehicles and autonomy stacks central to a military network; NODA is betting that the Pentagon will also pay for an independent layer capable of coordinating them.
Its role on a team led by established defense suppliers gives Duong and Corbett a route into a large federal prototype program. The amount NODA might receive has not been publicly established. Securing and delivering an award would require operational knowledge, working integrations and the ability to navigate a procurement system that often moves more slowly than the technology it is buying.