Case Studies

Generative AI Logistics for Communication and Execution (GAILFORCE)

Mission-Critical Analysis Automation

Generative AI Logistics for Communication and Execution (GAILFORCE)

A Generative AI-powered platform that enables users to create and modify complex, high-fidelity training scenarios for the Next-Generation Training System (NGTS) using natural language, significantly reducing development time and expertise requirements.
PROBLEM

Live, Virtual, Constructive (LVC) simulators like NGTS are essential for modern warfighter training and tactic development

However, these simulators are complex, often requiring weeks of dedicated training to create even simple scenarios. Scenarios quickly lose training value once repeated, making it difficult to effectively reuse scenarios for the same user. Furthermore, analyzing simulation results requires a separate set of specialized skills and cumbersome software, creating significant barriers to the effective and efficient use of these valuable training tools.

Knexus Solutions

SOLUTION
Knexus developed NGTS-GPT, a scalable, integrated Generative AI solution designed to remove barriers to the effective use of complex simulators like NGTS without requiring extensive specialized training. It leverages a fully functional prototype from Phase I and the proprietary KnexGPT framework to ensure security and model flexibility.

NGTS-GPT’s robust architecture is designed to:

Natural Language Scenario Generation

Allow users to generate, modify, and analyze scenarios by simply speaking or typing instructions, including complex mission-level concepts (e.g., waypoints, engagement lanes, event-based tactics).

Platform-AI Specification

Enable users to define highly specific, dynamic AI behaviors for individual platforms or groups (moving beyond pre-defined tactics) using natural language.

Advanced Analysis

Establish a direct, tight integration with the NGTS Analysis and Reporting Tool (ARTool) database, enabling users to ask any question about the simulation data in natural language, receiving tailored answers and dynamic visualizations.

Automated Communication Assessment

Ability to analyze pilot voice communication during scenarios in real-time, providing objective, quantitative metrics on brevity, accuracy, protocol adherence, and timeliness.

Security & Agnosticism

Utilize the KnexGPT framework to ensure the system is LLM-agnostic and can operate securely in classified environments (Secret and above), including deployment on JWCC Community Cloud or on-premises solutions.

Customer Outcome

CUSTOMER OUTCOME

NGTS-GPT dramatically reduces the learning curve and man-hours required for scenario creation, making high-fidelity LVC training accessible to a broader range of personnel.

By enabling natural language control over AI behavior, it allows instructors to rapidly create complex, non-deterministic, and tailored training events, leading to better-prepared warfighters. The integrated analysis capabilities, including the debrief dashboard and automated audio measures, ensure that every simulation run yields deeper, more actionable insights. The system is designed for secure, classified operation, maximizing its potential for widespread deployment across NAVAIR and the broader DoD.
GOOGLE TECHNOLOGY UTILIZED

Generative AI Logistics for Communication and Execution (GAILFORCE)

A Generative AI-powered platform that enables users to create and modify complex, high-fidelity training scenarios for the Next-Generation Training System (NGTS) using natural language, significantly reducing development time and expertise requirements.
Generative AI Logistics for Communication and Execution (GAILFORCE)
PROBLEM

The Defense Logistics Agency (DLA) is responsible for managing a vast and interconnected global logistics network, which generates a significant amount of data from both internal systems and various external sources.

To achieve its mission objectives, the DLA needs timely and robust insights from this data flow. However, identifying, classifying, and summarizing this information to create a single, concise Situation Report (SITREP) is a skill-intensive and time-consuming manual task for DLA experts, and creates a bottleneck to timely dissemination of information.

Knexus
Solution

Knexus developed GAILFORCE, a scalable, end-to-end Generative AI platform that automates and refines the SITREP generation process. This solution transforms the initial challenge into a personalized, daily information feed of Automated SITREPs.

GAILFORCE uses a sophisticated, multi-step architecture to:

Ingest Data:  Import diverse data types and formats rapidly (e.g., internal documents, intelligence reports, new articles).

Filter and Analyze: Automatically clean, filter, remove duplicates, and assess data for relevance to the DLA mission.

Generate Insight: Leverage Retrieval-Augmented Generation (RAG) to align analysis with DLA policy and plans, highlighting known or potential impacts to the DLA.

Deliver Personalized Reports: Provide a high-level summary of relevant topics with an interactive, human-language feature, delivered as a personalized, daily information feed.

SITREP Queries: Allows users to ask natural language questions about the generated SITREPs and source material, providing a customized, conversational means for users to better understand the information.

Customer Outcome

Customer Outcome

Generative AI Logistics for Communication and Execution (GAILFORCE)

The GAILFORCE system automates the labor-intensive tasks of sorting and analyzing vast amounts of data, providing personnel with significant efficiency gains and a competitive edge in decision-making.

It quickly delivers mission-critical, actionable intelligence through a user-validated, personalized feed, allowing experts to concentrate on strategic analysis instead of manual reporting. Built on a modular and extensible architecture that follows zero-trust principles, this solution is designed to be scalable across the organization and can be securely integrated with existing tools, ensuring long-term operational effectiveness.

Additional Case Studies