Develop and implement a user-friendly, multi-Generative Artificial Intelligence (AI) model platform for Texas A&M System members.

The project will take place in a phased approach.

Phase 1: Requirement Analysis and Design (Summer 2024)

An ongoing, iterative requirements gathering and design process that involved meeting with stakeholders to understand their needs and expectations, defining the project's goals and success criteria, creating a detailed project plan and developing design specifications and architecture diagrams.

Phase 2: Application Development (Summer/Fall 2024)

Application Development included setting up the development environment, integrating essential generative large language models, building user interfaces, implementing token-based usage tracking and conducting initial testing and reviews.

Phase 3: Infrastructure and Implementation Development (Spring 2025)

In this phase, the project focused on deploying dev, staging, preview and production environments and the capability to create multiple instances of the new application and related components. Using Infrastructure as Code techniques, these environments and instances will be created to ensure standardization, consistency and scalability. This phase also created the necessary configuration to ensure secure data access and adherence to enterprise standards for data security and compliance.

Phase 4: Testing and Quality Assurance (Spring/Summer 2025)

This phase will consist of conducting various tests such as functional, performance, security and customer experience testing. We will identify and fix bugs, optimize the application’s performance, and perform customer acceptance testing (beta testing) with selected users. Additionally, all test cases and results will be thoroughly documented.

Phase 5: Deployment and Go-Live (Fall 2025)

The last phase focuses on verifying the application’s functionality, performance, and security. During this phase, we will identify and fix bugs, ensure optimal performance, and finish testing with selected users. Comprehensive documentation of all test cases and results will also be performed to maintain transparency and facilitate future reference.