The Department of the Air Force’s operational advantage depends on the performance and agility of its Defense Business Systems (DBS). Yet many of these systems carry decades of technical debt, slowing mission execution and consuming valuable human capital.

This session will cover our current process and delivers a live demonstration of a secure, agentic, Human‑in‑the‑Loop (HITL) AI greenfield modernization workflow purpose‑built for PEO BES. Attendees will see how AI agents rapidly reverse‑engineer legacy business codebases, extract functional knowledge, and generate modern, secure, cloud‑ready applications—compressing modernization timelines from years to months and reducing manual reverse‑engineering labor by hundreds of hours per system.

By automating DBS modernization, this approach directly supports the DAF’s enterprise optimization goals, accelerates delivery of mission‑critical capabilities, and frees Airmen and Guardians to focus on higher‑value operational tasks. The session concludes with an interactive Q&A and a real‑time demonstration of the multi‑agent architecture.

Strategic Context

The Secretary of War’s memorandum directs the Department to accelerate America’s military AI dominance by eliminating bureaucratic barriers, unleashing experimentation, and executing seven “Pace‑Setting Projects.”

This session directly aligns with those directives by showcasing an AI‑first modernization pipeline that transforms Defense Business Systems into secure, cloud‑native, Zero Trust‑aligned applications at unprecedented speed.

Modernization Pipeline Overview

1. AI‑Driven Reverse Engineering of Defense Business Systems

Action: Deploy specialized AI agents to ingest and reverse‑engineer legacy business codebases, capturing:

• Functional & non‑functional requirements
• Data models
• Workflows
• Architecture patterns

Engineering Guardrails:

• Deterministic extraction workflows
• Isolation of core business logic from obsolete or spaghetti code

HITL Oversight & Approval: System Subject Matter Experts (SMEs) and engineers review the AI-extracted data models and workflows to verify accuracy before authorizing the creation of the baseline.

Strategic Alignment: Targets DBS technical debt and supports the DAF AI Strategy for Enterprise Optimization & Workforce Augmentation.

Measurable Impact:

• Reduces initial baseline capture from months to hours
• Cuts manual analysis labor by 70–90%

2. Human‑in‑the‑Loop Validation & Modernization Triage

Action: Provide interactive dashboards for PEO BES reviewers to validate extracted artifacts and determine modernization readiness.

Engineering Guardrails:

• Explainability tooling enabling traceability of AI‑generated artifacts
• HITL approval gates

HITL Oversight & Approval: This phase serves as a dedicated human approval gate. Program leadership must explicitly sign off on the auditable baseline before any modernization efforts begin.

Strategic Alignment: Reinforces Air Force Human‑Machine Teaming doctrine and ensures the DAF truly Owns the Technical Baseline.

Measurable Impact:

• Establishes authoritative, auditable baselines for every system
• Prevents legacy defects from propagating into modernized systems

3. Agile Requirement Curation (“Clean the Slate”)

Action: For systems selected for modernization, program offices refine the captured baseline—removing outdated requirements, merging new capabilities, and defining greenfield specifications.

Engineering Guardrails:

• ModelOps discipline
• Version‑controlled schemas and requirement sets

HITL Oversight & Approval: Program Managers and Functional Leads manually curate the requirement backlog and must formally approve the finalized "greenfield specifications" before code generation is allowed.

Strategic Alignment: Ensures alignment with DoD Business Enterprise Architecture and IT Enterprise Architecture, eliminating inherited technical debt.

Measurable Impact:

• Produces clean, modern requirement sets ready for automated generation
• Reduces requirement churn and rework during development

4. Zero‑Debt Greenfield Generation of Modern Applications

Action: Use a multi‑agent orchestrated pipeline to generate modern, secure, cloud‑native applications from validated specifications.

Engineering Guardrails:

• Modern Software Practices: Automated DevSecOps, CI/CD, SBOM generation, SAST/DAST
• Industry Standards: SOLID principles, secure coding patterns
• Cloud‑Native: Containerized, microservices‑based architectures
• Zero Trust & API‑First: ICAM‑enforced access, enterprise‑aligned APIs
• AI TRiSM: Runtime shields, telemetry, drift monitoring

HITL Oversight & Approval: DevSecOps teams and Cybersecurity Professionals conduct code reviews, evaluate SAST/DAST results, and provide final human authorization before deploying the generated code to production environments.

Strategic Alignment: Applications deploy directly into DAF enterprise environments (Platform One, Cloud One), accelerating delivery for PEO BES.

Measurable Impact:

• Generates production‑ready code in days instead of months
• Ensures RMF‑aligned artifacts and SBOMs are produced automatically

5. Continuous Sustainment via AI Feedback Loop

Action: Maintain the modernized system through continuous AI‑assisted updates—bug fixes, feature requests, and security patches.

Engineering Guardrails:

• Same DevSecOps, Zero Trust, and TRiSM controls as initial build

HITL Oversight & Approval: Human operators and maintainers review, test, and approve all AI-recommended patches and feature updates before they are merged into the live system.

Strategic Alignment: Supports the DoD AI Cybersecurity Risk Management Tailoring Guide and enables Continuous ATO (cATO).

Measurable Impact:

• Sustainment cycles shrink from weeks to hours
• Ensures continuous compliance and security posture

Who Should Attend

• Program Managers and Functional Leads
• Enterprise Architects and Solution Architects
• Cybersecurity and RMF Professionals
• Acquisition and Contracting Personnel
• Software Engineers and DevSecOps Teams

Interactive Demonstration & Q&A

Attendees will see the agentic pipeline in action—from legacy code ingestion to greenfield application generation—followed by an open Q&A.