Shared Responsibility Model

The Shared Responsibility Model outlines the division of responsibilities among Trusst AI, its customers, and AWS. This ensures clarity in managing and operating the Trusst AI platform deployed within

Trusst AI Responsibilities

Trusst AI is responsible for the application layer of its platform, ensuring the solution operates as expected, adheres to security best practices, and aligns with customer use cases. Specific responsibilities include:

1. Application Performance and Functionality

• Delivering a robust, secure, and functional platform aligned with customer objectives.

• Conduct testing, including end-to-end testing, prior to any deployment.

• Designing and maintaining deployment processes (CloudFormation templates, CDK scripts) for efficient and secure deployment into customer AWS environments.

• Ensuring the infrastructure aligns with AWS best practices and cost optimisation principles.

• Minimising access privileges in IAM roles required for deployment pipelines.

1. Security Best Practices

• Enforcing high-security standards across all deployments:

• Proactively identifying, testing, and addressing vulnerabilities and risks during the software development life cycle to ensure the platform’s security and reliability prior to deployment. In addition, any newly discovered vulnerabilities or risks that arise post-deployment will be resolved in accordance with the defined Service Level Objectives, ensuring continuous protection and operational stability.

• Continuous security monitoring and vulnerability scanning of Trusst AI owned environments running Trusst AI’s latest release versions to identify potential vulnerabilities.

1. Cost Optimisation

• Proactively optimise infrastructure and platform to optimise infrastructure usage.

• Implementing efficient database settings, auto-scaling, and resource fine-tuning.

• Ensuring Lambda functions are optimised through reserved concurrency, throttling, and other configurations.

1. Platform Updates

Trusst AI takes a structured approach to developing, testing, and deploying platform updates to ensure security and performance:

Proactive Development and Testing

• Developing features based on customer feedback, industry trends, and technological advancements.

• Conducting end-to-end testing, including security assessments, prior to making updates available.

Customer Control

• Publishing detailed release notes for transparency on features, fixes, and improvements.

• Allowing customers to opt-in via the Trusst Support Portal.

Secure Deployment

• Deploying updates securely using Trusst AI’s deployment pipelines.

• Ensuring all deployments adhere to security and operational best practices.

Types of Updates

• Feature Enhancements: New functionalities to improve platform capabilities or address specific needs.

• Optimisation Updates: Adjustments to platform components (e.g., model fine-tuning, Lambda runtime updates) to enhance performance and cost efficiency.

1. Support and Incident Management

• Resolving platform-related incidents within the defined Service Level Objectives.

• Providing customers with detailed release notes for new updates.

• Providing support via the Trusst Support Portal.

Customer Responsibilities

The customer is responsible for managing their AWS environment, ensuring the smooth operation of Trusst AI’s platform, and overseeing operational costs. Key responsibilities include:

1. AWS Account Management

• Provisioning, maintaining, and operating the AWS accounts where Trusst AI’s platform is deployed.

• Providing accurate network-specific parameters during setup, such as VPC CIDR ranges, unless using default Trusst managed network configurations.

• Provisioning the Trusst IAM Role for deployment processes.

1. Update Management

• Reviewing release notes provided by Trusst AI for platform updates.

• Opting into updates and ensuring that the Trusst AI access role is re-enabled (if previously disabled) to allow deployment pipelines to apply updates.

1. Financial Responsibilities

• Paying all AWS costs incurred by the operation of Trusst AI’s platform in the customer’s AWS accounts.

• Preventing misuse or excessive use of the platform for non-approved activities that could lead to unexpected costs, e.g. ensuring appropriate user role assignment for Role Based Access control.

• Implement appropriate AWS cost saving mechanisms such as AWS Savings Plans, Enterprise Discount Program, etc. e.g. Reserved Instances or spot instances for cost savings.

• Provide feedback on platform usage patterns or participating in periodic usage reviews with Trusst AI.

1. Platform Monitoring and Feedback

• Monitoring platform performance to confirm it performs as expected.

• Reporting issues promptly via the Trusst AI Support Portal.

1. Collaboration with AWS

• Engaging AWS to resolve issues or requests outside the scope of Trusst AI or the customer, such as: Service quota increases - refer to AWS Shared Responsibility Model.

• Enable/manage resolution of AWS-specific incidents.

AWS Responsibilities

AWS plays a crucial role in providing the infrastructure and support required to operate Trusst AI’s platform effectively. Refer to the AWS Shared Responsibility Model for details related to AWS Roles and Responsibilities related to your AWS Cloud infrastructure on which Trusst AI’s platform runs, e.g. availability of services required by Trusst AI’s platform.

Conclusion

This Shared Responsibility Model ensures a collaborative approach to managing and operating the Trusst AI platform. Trusst AI focuses on delivering a secure, cost-efficient, and high-performing application layer, while customers manage their AWS environment and operational costs. AWS supports the underlying infrastructure, creating a seamless, secure, and efficient ecosystem for Trusst AI’s solutions.

Summary Table

Here is a concise table summarising the responsibilities of Trusst AI, customers, and AWS based on the above shared responsibility model:

Area	Trusst AI	Customer	AWS
Platform Deployment	- Develop and maintain secure deployment pipelines.	- Provision AWS accounts and IAM roles for deployment.	- Provide infrastructure for secure and reliable deployment.
Application Performance	- Ensure the platform operates as expected.	- Monitor platform performance and report issues promptly.	- Ensure service availability and infrastructure reliability.
Security	- Proactively identify, test, and address vulnerabilities during development. - Resolve post-deployment vulnerabilities per SLOs. - Adhere to security best practices, including encryption, KMS, and regular assessments.	- Prevent misuse or unauthorised activities. - Report identified security risks promptly.	- Maintain physical security and service-level compliance of the underlying infrastructure.
Platform Updates	- Develop feature enhancements and optimisation updates. - Publish detailed release notes. - Deploy updates securely to customer environments (post opt-in).	- Review release notes and opt-in for updates via the support portal.	- Ensure compatibility of AWS services for updates.
Cost Optimisation	- Optimise platform components (e.g., auto-scaling, Lambda settings).	- Pay AWS costs related to platform operation. - Prevent excessive use or misuse of the platform.	- Offer cost-saving features (e.g., Reserved Instances, Savings Plans).
Support and Incident Management	- Resolve incidents within SLO timelines. - Provide detailed updates and escalation paths for critical incidents.	- Report incidents through the support portal. - Collaborate in incident resolution.	- Resolve AWS-specific incidents (e.g., quota increases, infrastructure issues).
Collaboration with AWS	- Support escalation to AWS when required.	- Engage AWS for resolution of AWS-specific issues beyond Trusst AI's scope.	- Provide timely support for AWS service-related issues.
Customer Onboarding	- Provide documentation and training to customers for effective platform usage.	- Complete onboarding activities, including understanding platform requirements and guidelines.	- Maintain documentation and tools for AWS services.
Financial Responsibilities	- Minimise operational costs through optimisation.	- Cover AWS operational costs. - Implement cost-saving mechanisms like Reserved Instances or Savings Plans.	- Ensure transparent billing and cost-saving programs for AWS services.