Various technical risks and technical debts were identified during the design of the EDDIE system. Technical risks refer to unwanted or unexpected behaviors that could negatively impact the system's operation. Technical debt encompasses identified behaviors, solutions, and concepts that could enhance system performance but have not been implemented.

Risks

The identified risks are categorized as either risks for developers or risks for users, as further discussed below.

Risks for Developers

Risk	Description	Applicable Measures
Monolithic Architecture	EDDIE is developed and deployed as a monolithic application primarily using Java. This architecture limits extensibility because extensions and plugins can only be developed in Java, preventing the use of other programming languages.	Gradually transition to a modular architecture or adopt a microservices approach to allow the integration of other languages. Additionally, define and implement a clear plugin API that supports cross-language compatibility through standard protocols (e.g., REST, gRPC).
Multitenancy Workarounds	EDDIE was not initially designed to support multitenancy. The current implementation supports multitenancy using workarounds, e.g., via additional parameters and processes, introducing potential risks related to security and stability.	Implement tenant isolation through database-level segregation or row-level security. Refactor the application to natively support multitenancy with scalable and secure tenant management capabilities.
Security	The lack of a robust, multitenancy-first design increases the risk of data breaches or unauthorized access.	Conduct regular security audits focusing on multitenancy-specific risks and implement tenant isolation mechanisms, such as sandboxing or scoped access controls, to enhance security.
Stability	Multitenancy workarounds may lead to unpredictable behavior or system crashes under specific conditions.	Perform thorough load testing and edge-case testing to identify and address stability issues. Develop a migration plan to refactor the system for native multitenancy support.

Risks for Users

Risk	Description	Applicable Measures
Unvalidated Data	In some cases, data sent by AIIDA might be unvalidated data. This data may diverge from the validated historical values that can be acquired from the Regional Data-sharing Infrastructures. This divergence can affect the accuracy of the output of the services provided by the eligible party.	Implement stricter data validation mechanisms at the point of entry to ensure all incoming data meets expected standards. Further develop fallback strategies or error handling to mitigate the impact of invalid data.
API Instability	The Regional Data-sharing Infrastructures may modify their APIs without prior notice, leading to errors in the Region Connectors of the EDDIE Framework. This can disrupt the historical data integration.	Update the affected Region Connector. Establish automated monitoring tools to detect API changes and alert developers. Maintain a buffer layer in the architecture that decouples direct API usage, enabling easier updates when changes occur.
Upload Bottlenecks	When AIIDA uses the local area network of the customer, sending high volumes of real-time data may cause network bottlenecks due to limited upload bandwidth.	Reduce the volume or frequency of near real-time data transmissions.

Technical Debt

The following areas represent unresolved technical debt that needs to be addressed to ensure normal system operation.

Debt	Description	Applicable Measures
API Change Detection	Related to the risk of API instability (previously discussed), there is no implemented mechanism to identify or anticipate changes in third-party APIs. A proactive solution, such as subscribing to relevant newsletters or implementing automated monitoring tools, should be developed to enable developers to adapt API calls in due time.	Develop a subscription process for API updates through newsletters or RSS feeds. Implement automated tools (e.g., API schema diff tools) to detect changes in third-party APIs and provide actionable insights to developers.