Skip to main content

Built for CGSC. Open to the Force.

Every Soldier
a Builder.

Making the Army the world's largest software developer, one agent at a time.

The Army's platforms already give you the tools. This site shows you how to use them, from your first AI agent to a deployed capability the force can scale.

Scroll
A soldier-developer is not a software engineer. Their role is to create practical problem-solving tools to address local unit and operational needs.
Read the foundational article

The Army has fielded powerful digital platforms, Army Vantage and GenAI.mil, that do something their designers didn't fully anticipate: they gave soldiers built-in tools to create their own software, including AI agents.

You don't need to be a software engineer or have a computer science degree to begin building. All you need is a problem worth solving and the willingness to learn how these platforms work.

This site exists to lower that bar, with real examples, real agents, and a community of builders doing exactly this at CGSC and across the force.

Built by Soldier-Developers.
Deployed. Real. Replicable.

Four AI agents, each named for a classical figure, built and deployed inside Army-approved platforms at CGSC. Each one is a proof of concept and a teaching example.

Athena
Grading Army Vantage

AI grading assistant that simulates a council of 50 faculty members to evaluate student essays.

  • ·Five grading archetypes from Rubric Hawk to Lenient Grader ensure full assessment spectrum
  • ·Reduces individual faculty bias through ensemble scoring
  • ·Within five percentage points of faculty grade 84 percent of the time
Ares
Wargaming Army Vantage

AI-enabled wargaming agent for Corps and Division-level exercises at CGSC.

  • ·Acts as blue team, red team cell, and white cell adjudicator simultaneously
  • ·Each course of action executed 100 times to find probabilistic outcomes
  • ·900 simulated turns vs. 2 completed by analog staff groups in the same exercise
Socrates
Assessment Army Vantage GenAI.mil

AI assessment agent that uses the Socratic Method to evaluate student mastery through dialogue.

  • ·Students must demonstrate Bloom's Taxonomy-aligned learning before their artifact is generated
  • ·Evaluator and Report Compiler sub-agents produce a faculty report and actionable student feedback
  • ·99 percent of students chose to keep learning after being offered their final grade
Aristotle
Pedagogy Army Vantage GenAI.mil

Free-form pedagogical agent that tutors and mentors students on course material before class.

  • ·Transforms static course readings into interactive, conversational learning experiences
  • ·Students engage with assigned texts on their own terms before walking into class
  • ·Supports stated learning objectives across multiple levels of learning
639
Total Users
2,080
Total Sessions
4
Deployed Agents

2025 to 2026 CGSC Academic Year

Built, Deployed, and Documented.

Every agent on this site has a published record. These articles trace the thinking, the building, and the results. From the foundational concept to operational agents running inside Army-approved platforms. The Soldier-Developer concept is not a theory, it is a bottom-up movement originating from CGSC.

War on the Rocks Mar 2026
Every Soldier a Software Builder: Governing the Army's New Digital Workforce

The foundational piece introducing and defining the Soldier-Developer concept. It proposes a three-level governance architecture, and argues the Army is already becoming one of the world's largest software organizations, whether it recognizes it or not.

Small Wars Journal May 2026
AI-First Professional Military Education: Validating the Grade Chain Before the Kill Chain

A case study of Athena and introducing the concept of Socrates. This article argues if PME won't trust AI to help grade essays, how can the Army trust AI in the kill chain? Agentic experimentation in PME is necessary to validate the hypothesis of the Department of War's AI-first future.

Small Wars Journal Jan 2026
AI-Enabled Wargaming at CGSC: Implications for PME and Operational Planning

A detailed account of the November 2025 CGSC exercise and the operational debut of Ares. Staff groups using analog methods completed an average of two turns. Staff groups using Ares completed up to nine turns, with each turn executed 100 times to find the probabilistic outcome of a course of action. 900 simulated turns versus two. This article documents what happened, what it means for PME, and what it suggests for operational planning.

Army.mil Jan 2026
CGSC Leads AI Integration in Professional Military Education

Official Army coverage of the Ares wargaming exercise and CGSC's broader AI integration efforts. In a little over a week, at zero cost, an AI-enabled wargame was created with career-influencing impacts for students. This marks the moment CGSC began transforming from just another schoolhouse into a no-code AI software development hub for the Army, empowering graduating officers to arrive at their new units as immediate change agents equipped with ready-to-use solutions.

Your first agent is closer than you think.

Building an AI agent doesn't require a computer science degree or any coding experience. Pick an approved platform, identify a problem your unit has, and build something real. Here's how it works.

Step 01
Choose Your Platform

Army Vantage and GenAI.mil are already approved and available to you. GenAI.mil is the faster starting point. Vantage is more powerful but has a steeper learning curve.

Step 02
Define Your Problem

The best agents solve a real problem someone actually has. Identify the repetitive task, the planning bottleneck, or the workflow your unit handles manually. That is your agent's mission.

Step 03
Build & Deploy

Use no-code environments to configure your agent, define its mission, and deploy it to your unit, all inside accredited and secure platforms you already have access to.

Start Building Today