Sometimes the underdog actually wins. Enabled Intelligence, an artificial intelligence company with a predominantly neurodiverse workforce, just landed the National Geospatial-Intelligence Agency's largest data labeling contract on Nov. 24, beating out a competitor backed by one of the world's biggest tech companies.

The victory is particularly striking because Enabled Intelligence went up against Scale AI, which received a $14 billion investment from Meta Platforms (META) back in June. The Sequoia contract is a single-award agreement with a ceiling value of up to $708 million over seven years, and it's a pretty big deal in the world of Pentagon AI.

What This Contract Actually Does

Think of data labeling as teaching computers to see. Before an AI system can automatically detect a tank in satellite imagery, humans need to trace thousands of tanks and label them so the algorithm learns what a tank looks like. That's what Enabled Intelligence will be doing for the National Geospatial-Intelligence Agency and the broader Department of Defense.

The contract supports data labeling for geospatial intelligence AI and machine learning capabilities across multiple programs within the intelligence community. The Falls Church, Virginia-based company will provide services that enable computer vision algorithms to perform automated object detection, object tracking, object classification, and pattern detection.

This isn't just academic work. The contract covers Maven, the military's primary targeting program, along with other flagship programs applying AI and machine learning to intelligence, surveillance, and reconnaissance sensors through computer vision. It also includes natural language processing, analytic models, and AI systems for business process automation.

The company specializes in transforming raw data into training information for mission-critical government and commercial applications, with particular expertise in sensitive and classified datasets.

The Neurodiverse Workforce Advantage

Here's where things get interesting. Approximately 60% of Enabled Intelligence's employees are on the autism spectrum, CEO Peter Kant told Bloomberg. The workforce has exploded from 33 employees to 136 people since the company learned of the contract award two months ago.

Kant said he developed this hiring approach after reading research indicating neurodiverse individuals possess unique capabilities for focused pattern recognition work. When your job involves meticulously tracing outlines of military equipment in satellite imagery for hours on end, attention to detail and pattern recognition aren't just nice to have—they're the entire game.

"Precision in support of American national security missions has been at the forefront of everything we do at Enabled Intelligence," Kant said in the company's statement. "Our open architecture approach, combined with our rigorously trained and skilled workforce, ensures that NGA and the broader DoD and Intelligence Community have access to the precise, unbiased data labeling necessary to deploy AI capabilities at the speed and scale."

Scale AI's Disappointment

Scale AI, which became a primary vendor for Project Maven in late 2020, was competing for the same contract. A company spokesperson acknowledged disappointment with the outcome, which makes sense when you're talking about a $708 million opportunity.

Scale AI isn't exactly hurting, though. The company maintains separate agreements with the U.S. Army and Pentagon valued at up to $199 million combined, and continues other defense work.

Former Scale AI co-founder and CEO Alexandr Wang, who has since joined Meta, wrote to President Donald Trump in January through letters and advertisements stating "America must win the AI War." Wang referenced the company's work for the Defense Department in his appeal to the president.

The Technical Details Matter

Kant prepared extensively before the contract decision by constructing secure facilities capable of handling classified information at secret and top-secret levels. This kind of infrastructure isn't cheap or easy to build, but it's absolutely necessary for the work.

The actual work involves tracing outlines of military equipment, including aircraft, armored vehicles, naval vessels, and personnel in satellite imagery using digital tools. This creates datasets for training computer systems to identify objects independently. It's painstaking work that requires both precision and the ability to maintain focus through repetitive tasks.

Enabled Intelligence said it will partner with BAE Systems, Vantor, Whiteboard Federal, and other firms on the program, bringing together a network of defense contractors to handle the classified AI and imaging work.

The win demonstrates that in government contracting, sometimes the combination of specialized workforce capabilities and proper infrastructure can beat out bigger names with deeper pockets.