AWS Public Sector Blog

Meeple, Padua University spin-off, launches Amazon Bedrock AI Plugin for Moodle

 

AWS branded background with text "Meeple, Padua University spin-off, launches Amazon Bedrock AI Plugin for Moodle"

Meeple Srl—a technology spin-off from the University of Padua and Amazon Web Services (AWS) Partner—has released the Amazon Bedrock API Provider plugin for Moodle. This open source integration enables educational institutions to seamlessly connect Moodle Learning Management Systems with Amazon Bedrock, providing access to foundation models (FMs) including Anthropic’s Claude. Initially developed to power the University of Padua’s Lucrez-IA virtual assistant project, the plugin is now available through the official Moodle Plugins directory and GitHub. The plugin offers a standards-compliant, General Data Protection Regulation (GDPR)-ready solution for educational institutions seeking to enhance teaching and learning experiences with generative AI while maintaining data sovereignty.

Lucrez-IA: A reference implementation at the University of Padua

The Amazon Bedrock API Provider plugin was born from the Lucrez-IA project at the University of Padua, Italy’s pioneering foray into institutional AI. Named in honor of Elena Lucrezia Cornaro Piscopia, the first woman to receive a university degree in the world (University of Padua, 1678), Lucrez-IA demonstrates the practical application of AWS generative AI in higher education at scale.

Serving one of Italy’s largest and oldest universities with over 70,000 students and 5,000 faculty and staff members, Lucrez-IA offers personalized AI assistance across three specialized bot interfaces: General Purpose, Digital Learning, and Document Analysis. The system has been implemented as a central institutional resource, with AWS infrastructure ensuring secure, GDPR-compliant operations entirely within European AWS Regions.

Along with to the web-based interface (accessible through Single Sign-On), Lucrez-IA is integrated directly into the university’s Moodle learning environments through custom blocks, making AI capabilities available at the point of teaching and learning. This integration has enabled faculty to more efficiently develop course materials, respond to student inquiries, and manage administrative tasks—allowing educators to focus more time on high-value interactions with students.

As the first public university in Italy to deploy a generative AI solution based on Anthropic’s Claude models, the University of Padua’s implementation serves as a benchmark for responsible institutional AI adoption in European education, balancing innovation with ethical considerations and regulatory compliance.

Benefits for students and educators

Beyond its technical elegance, the plugin delivers tangible value in the daily lives of educators and learners. Here are just a few real-world scenarios:

  • A professor quickly generates reading comprehension exercises using Claude’s text generation, freeing up hours normally spent editing materials.
  • A lecturer uses the summarization feature to condense complex research articles into digestible abstracts for undergraduates.
  • A student with dyslexia benefits from simplified summaries and voice-compatible content, enhancing accessibility and inclusion.

These examples highlight how the plugin doesn’t just add AI to Moodle—it brings smarter, faster, and more personalized learning experiences to the entire academic community.

Key capabilities of the Amazon Bedrock API provider for Moodle

The open source plugin developed by Meeple Srl uses Moodle’s AI subsystem (Moodle 4.5+) to provide access to Amazon Bedrock FMs through three core functionalities:

  • Text generation: Faculty can use Anthropic’s Claude models to create high-quality content for educational purposes. With customizable system instructions, this feature streamlines the development of learning materials, assessments, and communications. The implementation intelligently handles different model versions, maintaining compatibility as new FMs emerge.
  • Image generation: Users can create AI-generated images through text prompts using models such as the Amazon Nova Canvas. With customizable dimensions, this feature enables images to be automatically processed and stored in Moodle’s file system with appropriate watermarking. This empowers educators to quickly create visual aids without graphic design expertise.
  • Text summarization: This feature efficiently condenses lengthy texts into concise summaries while preserving key information. This capability is particularly useful for distilling research papers, creating abstracts, or summarizing case studies for student consumption.

Each capability is designed with education-specific use cases in mind and includes configurable model selection, allowing institutions to choose optimal FMs while managing costs effectively.

Text and image generation features are accessible directly through Moodle’s Rich Text Editor (TinyMCE), while text summarization can be used within course activities and content areas by teachers.

Figure 1: Text summarization

Architecture overview: How Moodle connects to Amazon Bedrock

The following architectural diagram illustrates how the Amazon Bedrock API Provider plugin connects Moodle’s AI subsystem to Amazon Bedrock.

The plugin acts as a bridge between Moodle and FMs hosted on AWS, such as Claude (Anthropic), Amazon Nova, and Cohere. Requests from Moodle (such as text generation, summarization, and image creation) are routed through the Bedrock Plugin, which authenticates and processes the input before passing it to Amazon Bedrock. The output is then returned and stored within Moodle’s ecosystem. This architecture operates entirely within EU-based AWS Regions.

Figure 2: Moodle – Amazon Bedrock Integration Architecture

Amazon Bedrock: Secure, compliant AI for education

The Amazon Bedrock API Provider uses the enterprise-grade security and compliance capabilities of AWS to address the specific requirements of educational institutions.

Data privacy and GDPR compliance

The plugin enables educational institutions to maintain full control over their data while accessing advanced AI capabilities. Using regional AWS infrastructure in Europe allows institutions to maintain GDPR compliance with data processing entirely within EU boundaries. The cross-Region inference capabilities with European Region prefixes (eu.*) enable data sovereignty while maintaining high performance and reliability.

Model selection flexibility

Institutions can choose from multiple FMs available through Amazon Bedrock, such as various Claude models optimized for different tasks and performance requirements. Amazon Bedrock provides access to models from leading AI companies and supports custom models that can be imported by institutions. This flexibility allows organizations to balance capability requirements with cost considerations, adapt to evolving AI technologies, and even deploy domain-specific models tailored to their educational needs.

Enterprise-grade security

AWS security credentials management, with appropriate permissions for Amazon Bedrock APIs, enables AI interactions to be properly authenticated and authorized. The plugin’s implementation includes comprehensive rate limiting at both user and global levels, allowing institutions to manage usage and costs effectively.

Figure 3: Plugin setup

Get started with AI in your educational environment

The Amazon Bedrock API Provider for Moodle offers educational institutions a direct path to implementing AI capabilities by following these steps:

  • Set up an AWS account with Amazon Bedrock access
  • Download the plugin from the official Moodle Plugins directory
  • Install the plugin in your Moodle 4.5+ environment
  • Configure your AWS credentials and choose your preferred FMs

This solution addresses key concerns in educational technology: responsible AI implementation, data privacy, and integration with existing teaching systems. The University of Padua’s Lucrez-IA project demonstrates how these tools can enhance administrative efficiency and teaching effectiveness.

The plugin is open source, with its code available on GitHub for community contributions and adaptations. Organizations can deploy the solution with various FMs according to their specific needs and policies.

We encourage educational institutions to explore how Amazon Bedrock and Moodle can work together to responsibly augment learning experiences while maintaining institutional values and requirements.

Dario Da Re

Dario Da Re

Dario has been the director of the Digital Learning and Multimedia Office at the University of Padova since 2017. With a PhD in Social Research Methodology, he coordinates innovative generative artificial intelligence projects for teaching, including the Lucrez-IA platform and its integrations into software and applications. He is responsible for the main university technological platforms (Moodle, Zoom, Kaltura) and is an adjunct professor of digital transformations in the Sociological Sciences degree program.

Angelo Calò

Angelo Calò

Angelo is head of digital learning at the University of Padova with over 15 years of experience in educational technology and AI integration. He coordinates Moodle platform development and online training services, focusing on artificial intelligence solutions for education. Together with colleagues Davide Ferro and Dario Da Re, he developed Lucrez-IA, an innovative AI system leveraging Amazon Bedrock and Anthropic Claude to enhance learning experiences.

Davide Ferro

Davide Ferro

Davide is a solutions architect at the University of Padova's Digital Learning and Multimedia division, specializing in AWS cloud architectures for educational technology. He focuses on integrating AI systems into learning platforms and supporting digital learning applications, including Moodle. Together with colleagues Angelo Calò and Dario Da Re, he developed Lucrez-IA, an innovative AI system leveraging Amazon Bedrock and Anthropic Claude to enhance learning experiences. Davide also serves as solutions architect at Meeple srl, a University of Padova spin-off.

Francesco Polimeni

Francesco Polimeni

Francesco is a senior solutions architect on the Global ISV team at AWS with over 4 years of experience helping software vendors innovate and scale their solutions globally. He brings more than 10 years of expertise in machine learning, making him particularly effective in guiding partners through their AI and ML transformation journeys. Francesco actively pursues bridging the gap between scientific research and innovation.

Andrea Catalano

Andrea Catalano

Andrea is a senior solutions architect at AWS with over 5 years of experience, specializing in the public sector with a particular focus on universities and research centers across Italy. He provides strategic cloud guidance to help customers architect their AWS environments for optimal cost efficiency, security, resiliency, and elasticity.