As we return to work after the holiday break, the Scale AI Public Sector team wanted to reflect on our work heading into 2025.&nbsp;
Over the past four years, we have been dedicated to strengthening U.S. national security by equipping U.S. agencies with the best-in-class commercial AI technologies. As strategic rivalries continue to intensify and adversaries form new alliances globally to challenge U.S. leadership in AI, the mission of Scale&rsquo;s Public Sector team has never been more vital. We are dedicated to ensuring that the U.S. and its allies have the best technology to lead in this increasingly complex global landscape.
The snapshot below captures a few key highlights from last year:
<ul>
<li dir="ltr" aria-level="1">
Scale deepened its ties to Missouri&rsquo;s geospatial and defense tech industries
</li>
<li dir="ltr" aria-level="1">
We developed novel capabilities with DoD innovation centers to create scalable solutions for perimeter security&nbsp;
</li>
<li dir="ltr" aria-level="1">
We partnered with leading defense-tech companies like Anduril to innovative start-ups like Holos that have built an Augmented Reality training platform
</li>
<li dir="ltr" aria-level="1">
We introduced a new product, Defense Llama, which is a Large Language Model (LLM) built on Meta&rsquo;s Llama 3 that is specifically customized and fine-tuned to support American national security missions
</li>
</ul>
St. Louis Commitment
As for some other highlights, Scale also began training the AI workforce of the future right here in the United States, specifically Missouri. To do so:&nbsp;
<ul>
<li dir="ltr" aria-level="1">
Scale deepened its ties to Missouri&rsquo;s geospatial and defense tech industries with our <a href="https://scale.com/blog/scale-stl-center?utm_source=linkedin&amp;utm_medium=organic-social&amp;utm_campaign=FY2410-BRAND-ORGSoc-stl-center">St. Louis AI Center</a>, which produced millions of high-quality geospatial data annotations. These annotations power the systems used by government leaders to make mission-critical decisions.&nbsp;
</li>
<li dir="ltr" aria-level="1">
Our St. Louis AI Center Lead <a href="https://sg001-harmony.sliq.net/00325/Harmony/en/PowerBrowser/PowerBrowserV2/20200831/-1/11596?mediaStartTime=20241029101143&amp;mediaEndTime=20241029105310&amp;viewMode=3&amp;globalStreamId=4">testified before</a> the Missouri State House of Representatives Special Interim Committee on Workforce Innovation and Technology on how to build an AI workforce.&nbsp;
</li>
<li dir="ltr" aria-level="1">
Also in St. Louis, Scale established partnerships with three local veteran-owned, minority-owned, or woman-owned small businesses in the local community, and we are proud to become members of <a href="https://greaterstlinc.com/">Greater St. Louis</a>, an organization committed to creating economic growth and jobs in St. Louis.
</li>
</ul>
Our Technology in Action
We partnered with the U.S. government and its innovation centers to develop and deploy new AI capabilities. Our flagship product Scale Data Engine underpinned a collaboration between the Defense Innovation Unit (DIU) and the United States Air Force (USAF) that provided the DoD novel capabilities to monitor large perimeter areas worldwide without external power sources. Scale Data Engine features, including data curation, data labeling, and model test and evaluation, enabled us to develop bespoke computer vision models for maritime vessel detection in a timely and cost-efficient manner. Other highlights include:&nbsp;
<ul>
<li dir="ltr" aria-level="1">
<a href="https://scale.com/blog/scale-partners-with-cdao-to-test-and-evaluate-llms">Partnering</a> with the Department of Defense&rsquo;s Chief Digital and Artificial Intelligence Office (CDAO) to develop a rigorous test and evaluation framework for LLMs within the DoD, and embarking on an enterprise-level AI scaffolding partnership with them to build the DoD&rsquo;s AI-ready data foundation. This will help enable rapid AI development and ensure that models are accurate and relevant to their missions.&nbsp;
</li>
<li dir="ltr" aria-level="1">
The Defense Information Systems Agency (DISA) selected Scale to execute the Project ASCEND Prototype Other Transaction Agreement (OTA). This agreement focuses on deploying cutting-edge Generative AI solutions tailored for Defensive Cyber Operations (DCO). The awarded prototype will assist DoD information networks (DODIN) defenders by processing and synthesizing massive amounts of critical data, enhancing cybersecurity, and national defense capabilities.
</li>
<li dir="ltr" aria-level="1">We have a contract with Defense Logistics Agency (DLA) Research and Development to use Scale Donovan to streamline its operations and create real-time data insights. DLA manages the U.S. military&rsquo;s entire supply chain, which last year alone delivered over $47 billion in goods and services via a massive global distribution network and roughly 25,000 employees. With Donovan, DLA will be able to automate routine tasks and improve decision-making processes to better support our military&rsquo;s logistics and acquisition communities.</li>
</ul>
Partnerships
As part of our commitment to working with the public sector more broadly, we have engaged in a number of partnerships and collaborations:&nbsp;
<ul>
<li dir="ltr" aria-level="1">
We are members of <a href="https://www.anduril.com/lattice-partner-program/">Anduril&rsquo;s Lattice Partner Program</a>. Combining Scale Donovan, with Lattice, Anduril&rsquo;s Command and Control software, creates an end-to-end solution that delivers a decision advantage through software and AI-enabled joint staff planning. The integration will enable users to access Lattice&rsquo;s orchestration, planning and simulation capabilities, and extend Donovan&rsquo;s AI capabilities to ensure true machine-speed command and control.&nbsp;
</li>
<li dir="ltr" aria-level="1">
Scale j<a href="https://scale.com/blog/scale-joins-us-ai-safety-consortium">oined</a> the U.S. Artificial Intelligence Safety Institute Consortium, collaborating with NIST and other leaders to set industry-wide AI safety standards and testing protocols.&nbsp;
</li>
<li dir="ltr" aria-level="1">
We are partnering with <a href="https://holos.io/government.html">Holos</a> to use AI to rapidly create Augmented Reality (AR) training for the U.S. Air Force (USAF). This capability will enable USAF instructors to create new AR training exercises in record time and rapidly create hundreds of modules that will ensure that our airmen are prepared for any situation they may encounter.
</li>
<li dir="ltr" aria-level="1">
<a href="https://scale.com/blog/scale-hosts-jcdcai-cyber-ttx">Scale also collaborated with the U.S. Cybersecurity and Infrastructure Security Agency (CISA)</a> to host the second Joint Cyber Defense Collaborative AI Cyber Tabletop Exercise in San Francisco. This event brought together more than 90 experts from government and industry to explore collective solutions for addressing&nbsp; AI security challenges.&nbsp;
</li>
</ul>
Capabilities
We launched new products and iterated on existing ones to bring our warfighters and civilian agencies the leading AI technologies. These capabilities are underpinned by our commitment to delivering secure, cutting-edge technology to the broader federal landscape. In addition to our Impact Level 4 (IL-4) Provisional Authorization from the Defense Information Systems Agency (DISA), we are proud to have earned <a href="https://scale.com/blog/scale-achieves-fedramp-high-authorization">FedRAMP&reg; HIGH authorization</a> and to be able to support classified environments (SIPR/JWICS) in DoD and IC settings.
<ul>
<li dir="ltr" aria-level="1">
Scale introduced <a href="https://scale.com/blog/defense-llama">Defense Llama</a>, a purpose-built LLM built on Meta&rsquo;s Llama 3 that is customized and fine-tuned to empower our service members and national security professionals to apply the generative AI to their unique use cases, such as planning military or intelligence operations and understanding adversary vulnerabilities.
</li>
<li dir="ltr" aria-level="1">
We launched<a href="https://scale.com/blog/responsible-ai-scale-evaluation-for-public-sector"> Scale Evaluation</a> and<a href="https://scale.com/leaderboard"> Leaderboards</a>, based on research from our Safety, Evaluations, and Alignment Lab (SEAL), so public sector partners can also measure AI performance, safety, and reliability using industry-leading benchmarks.&nbsp;
</li>
<li dir="ltr" aria-level="1">
<a href="https://scale.com/blog/celebrating-one-year-of-scale-donovan">Scale Donovan </a>has now been out for over a year and now integrates LLM-based tools like geospatial chat, text-to-API, and retrieval-augmented generation, to enable public sector teams to swiftly access verified intelligence and reduce mission planning time from hours to minutes.&nbsp;&nbsp;
</li>
</ul>
Our long history of serving the United States as a partner for AI development will continue in 2025. From the Public Sector team at Scale, <a href="https://scale.com/blog/veterans-day-2024">which includes dozens of Veterans</a>, we wish everyone a Happy New Year and look forward to continuing our unwavering support of the United States of America.&nbsp;

The Scale AI Public Sector team wanted to reflect on our work heading into 2025. 