.svg)
A complex dataset with seven distinct item types needed comprehensive cleaning and parsing. The project required a deep understanding of the unique attributes of STAR exam items. Additionally, a model was necessary to grasp the nuances across various item types, grade levels, and skill sets. Lastly, maintaining observability, governance, and compliance throughout the AI-powered content creation process posed a significant challenge.
Content developers could benefit from a simplified process for prompt creation. This project offered the chance to revolutionize how exam items for the STAR exam are produced across various grade levels and skills. It also aimed to empower educators with AI-powered tools to enhance their creativity and efficiency. Incorporating observability, governance, and compliance measures was essential to ensure responsible and ethical AI use for educational content development
Renaissance Learning leveraged Amazon Web Services (AWS) to fine-tune a large language model (LLM) for educational content development. This collaboration used AWS Jumpstart to iterate through various prompt designs and configurations for optimal performance. As the project evolved, the team transitioned to the 70 billion parameter Llama2 model, utilizing AWS Jumpstart for enhanced performance capabilities. The fine-tuned model was then deployed on the Renaissance AVM platform for real-world applications.
To ensure responsible use of AI, AVM integrated observability tools to monitor model performance, established governance frameworks for ethical AI practices, and applied compliance measures to meet educational content standards.
This solution significantly reduces the time and effort required to create educational content. It streamlines the prompt creation process, making the technology both accessible and practical for content developers. The platform can generate complete exam items that cover a wide range of item types, grade levels, and skill sets. Content creators benefit from a user-friendly tool that responds to simple yet effective prompts. Additionally, the AI-powered content creation process is designed to be observable, governed, and compliant with educational standards, ensuring that it meets all necessary requirements.
This solution enhances the efficiency and creativity of content developers at Renaissance Learning. It facilitates the creation of high-quality educational content more quickly and easily, fostering a collaborative environment between AVM and Renaissance Learning that drives innovation in educational technology. This partnership positions Renaissance Learning as a leader in AI-powered educational content development, laying the foundation for future breakthroughs that could transform the industry.
AWS Jumpstart's large language models (LLMs) protect intellectual property, ensuring competitors do not benefit from proprietary data. The observability, governance, and compliance measures provided by AVM underscore Renaissance Learning's commitment to responsible AI use.
.svg)