AI-Powered Analytics with Automated Cloud Operations on AWS
Industry: Education, Public Sector, Data Analytics
Platform: Amazon Web Services (AWS)
Client: CampusIL
Services: Generative AI, Data Analytics, Cloud Operations, DevOps Automation, Security, Application Integration
About the Customer
CampusIL is Israel’s national digital learning platform, led by the Israel National Digital Agency. It provides accessible online education and learning resources to millions of learners, educators, and organizations.
The platform generates significant volumes of educational data, including student engagement, course activity, and content interactions.
CampusIL uses this information to understand learner behavior, evaluate course effectiveness, and support data-driven decisions that improve educational outcomes.
Customer Challenge
CampusIL operated an advanced Business Intelligence environment based on a customized implementation of OpenEdX Aspects, using ClickHouse to store learning events and Apache Superset to provide analytics dashboards and visualizations.
Although the platform offered powerful analytical capabilities, creating new dashboards and queries required specialized SQL and ClickHouse knowledge.
This created ongoing dependency on technical experts for routine data exploration and dashboard development, increasing operational effort and slowing access to insights.
The large volume of educational events also made data exploration and query creation more complex.
CampusIL wanted to simplify the analytics experience by allowing users to request information and generate visualizations through natural language.
However, introducing generative AI created additional operational requirements. The AI system needed accurate, continuously updated knowledge of the underlying database structure, while maintaining secure access to educational data and supporting reliable production operation.
Manual updates to database metadata and AI context would have increased maintenance requirements and created the risk of outdated information being used for query generation.
CampusIL therefore needed a secure, automated, and efficiently operated analytics solution that would reduce dependency on specialized technical resources and simplify the ongoing maintenance of its AI capabilities.
Partner Solution
Peax Cloud designed and deployed an AI-powered analytics solution integrating Amazon Bedrock with CampusIL’s existing Apache Superset and ClickHouse environment.
The solution, deployed to production in July 2025, combines generative AI capabilities with automated data and application operations.
The implementation includes:
- AI-Powered Analytics: Amazon Bedrock interprets natural-language requests and generates ClickHouse queries and Apache Superset charts, allowing users to explore data without writing SQL.
- Application Integration: A React frontend and Python backend were developed and integrated into Apache Superset, enabling users to interact with the AI assistant directly within the existing BI platform.
- Automated RAG Pipeline: A cron-based Retrieval-Augmented Generation process updates the AI knowledge base with the latest ClickHouse schemas and metadata, maintaining relevant context for query generation.
- Security and Data Governance: Role-based access controls and secure communication between solution components help protect educational information and maintain appropriate data access.
- Data Privacy: The architecture was designed to keep data processing within the AWS environment and protect sensitive educational information.
- Operational Automation: Automated schema ingestion and knowledge-base updates reduce the need for manual maintenance and support consistent AI-assisted analytics operations.
Peax Cloud provides the cloud and DevOps expertise supporting the solution’s operation and evolution, with automated processes helping maintain the AI assistant’s awareness of the underlying analytics environment.
The automated RAG pipeline is a central operational capability. As database schemas evolve, the scheduled process updates the AI’s metadata knowledge, reducing the manual effort required to keep the solution aligned with the underlying data platform.
This combination of application integration, secure cloud architecture, and operational automation helps CampusIL operate a sophisticated AI analytics solution with reduced maintenance complexity.
Results and Benefits
Reduced Dependency on Technical Specialists: Users can generate analytics queries and charts through natural language, reducing reliance on BI developers for routine data exploration.
Reduced Operational Maintenance: Automated schema ingestion and RAG updates minimize manual knowledge-base maintenance and help the AI assistant remain aligned with database changes.
Faster Access to Insights: The conversational interface enables non-technical users to request analytics and visualizations without developing SQL queries.
Improved Operational Consistency: Automated processes provide a structured approach to maintaining AI metadata and supporting the continued operation of the analytics solution.
Secure Analytics Access: Role-based access controls and secure AWS integration support educational data governance and privacy requirements.
Improved Business Efficiency: CampusIL’s administrators and educational teams can spend less time on technical query development and more time interpreting insights and improving educational offerings.
Foundation for Future Innovation: The AWS-based architecture supports future enhancements such as multilingual analytics, automated dashboard recommendations, and additional AI-assisted capabilities.
By combining generative AI with automated operational processes, Peax Cloud helps CampusIL maintain an accessible and more efficiently operated analytics environment that supports its mission to improve digital education through data-driven insights.