ChurnShift Privacy Policy
OCT 5, 2026
Table of contents
1. Overview
2. Scope of the Policy
3. Data Collection and Processing
4. Access Control and Identity Management
5. Security, Encryption, and Architecture
6. AWS Well-Architected Architecture and Foundational Technical Review
7. AI, LLM, and Model Privacy
8. Data Retention and Deletion
9. Compliance and Regulatory Support
10. Changes to This Policy
11. AI-Assisted Drafting and Human Review Disclaimer
12. Contact Us
1. Overview
ChurnShift is committed to protecting the privacy, security, ownership, and control of our customers’ data and AI assets. Our approach is reflected in the ChurnShift motto:
“Your Data, Your Models, and Your Controls.”
This Privacy Policy explains how ChurnShift processes and protects information when the ChurnShift solution is deployed within a customer-controlled private cloud environment, including a customer’s Amazon Web Services (AWS) Virtual Private Cloud (VPC).
ChurnShift is designed so that customer data remains within the customer’s private cloud environment. Customers retain control over their data, models, access policies, encryption keys, and infrastructure.
This principle extends beyond the underlying customer data. AI and machine-learning workloads-including supported Large Language Model (LLM) services, inference services, and subsequent models trained or fine-tuned using customer data or customer-provided expertise-are designed to be deployed and served within the customer’s private cloud environment.
Accordingly, ChurnShift’s privacy architecture is based on three fundamental principles:
• Your Data: Customer data remains within the customer-controlled private cloud environment and is encrypted by default.
• Your Models: Models trained or fine-tuned using customer data or expertise remain deployed and served within the private cloud environment.
• Your Controls: Customers maintain control over access, identity management, encryption keys, security policies, retention, and infrastructure configuration.
2. Scope of the Policy
This policy applies to:
• All instances of ChurnShift software deployed in a customer-managed private cloud or VPC.
• All data processing activities carried out by the ChurnShift solution within that environment.
• Customer databases, database fields, object storage, including Amazon S3, and other supported storage services.
• AI and machine-learning workloads deployed as part of the ChurnShift solution.
• LLM services and other supported AI services integrated into the ChurnShift environment.
• Models subsequently trained, fine-tuned, customized, or otherwise created using customer data or customer-provided information within the ChurnShift environment.
3. Data Collection and Processing
3.1 Data Processed
ChurnShift processes only the data that customers choose to provide for authorized analytics, AI, machine-learning, churn prediction, customer-retention, and related purposes.
Depending on the customer’s use case, this may include:
Customer profile information, such as account details and engagement metrics.
Transactional data, such as purchase history and subscription events.
Behavioral data, such as product usage logs and interaction patterns.
Customer-provided documents, knowledge, or other information used by authorized AI or LLM functionality.
All such processing is designed to occur within the customer’s private cloud environment.
3.2 No External Data Transfer
ChurnShift’s architecture is designed around data residency within the customer’s private cloud environment. In particular:
ChurnShift does not export raw customer data to ChurnShift-managed external infrastructure.
ChurnShift does not store customer production data on ChurnShift-managed servers outside the customer’s environment.
Model training, fine-tuning, inference, analytics, and related processing are designed to occur within the customer’s private cloud environment.
Supported LLM services used by the ChurnShift solution are designed to maintain processing within the customer’s private cloud architecture.
Models subsequently trained or fine-tuned using customer data or customer-provided expertise are designed to remain deployed and served within the customer’s private cloud environment.
The objective of this architecture is that customer data does not need to leave the customer’s private cloud in order to obtain AI, machine-learning, analytics, or LLM capabilities provided through ChurnShift. This architecture implements ChurnShift’s principle:
“Your Data, Your Models, and Your Controls.”
4. Access Control and Identity Management
All access to customer data and ChurnShift resources is governed by the customer’s organizational access policies.
ChurnShift personnel do not have access to customer production data unless such access is explicitly authorized by the customer for purposes such as troubleshooting or support and is performed under customer authorization and supervision.
Administrative credentials and access permissions remain under customer control.
ChurnShift supports Role-Based Access Control (RBAC) so that access to applications, data, models, and administrative functionality can be restricted according to organizational roles and permissions.
ChurnShift user management is built on Amazon Cognito and supports integration with SAML-based identity providers, allowing organizations to integrate their existing enterprise identity-management infrastructure, including directory and identity systems such as LDAP- or Active Directory-backed identity providers, where configured through a compatible SAML identity provider.
This architecture enables customers to maintain centralized identity governance and apply their organization’s authentication, authorization, and access-management policies to ChurnShift.
5. Security, Encryption, and Architecture
ChurnShift follows security and cloud-architecture best practices and is designed around the principle that protection applies not only to files, but to customer information throughout the data and AI infrastructure. Security measures include:
Deployment in isolated, customer-controlled network environments.
Encryption in transit using TLS.
Encryption at rest by default for customer data.
Encryption of database data, including stored database fields, using AWS encryption capabilities and customer-managed encryption keys (CMEK) where provided by the ChurnShift reference architecture.
Encryption of object storage, including Amazon S3 storage, using customer-managed AWS encryption keys.
Customer control over applicable AWS Key Management Service (AWS KMS) keys.
Role-Based Access Control (RBAC).
Audit logging.
Customer-controlled identity and access-management policies.
Private-cloud deployment of customer AI and machine-learning workloads.
Private-cloud serving of supported LLM services and subsequent customer-trained or fine-tuned models.
5.1 Customer-Managed Encryption
ChurnShift’s AWS reference architecture is designed so that customer data at rest is encrypted by default using customer-managed encryption keys, including supported database and storage resources. This includes customer information stored in database services and object-storage services such as Amazon S3. Customer-managed encryption provides an additional layer of customer control because encryption-key administration remains within the customer’s AWS environment and can be governed through the customer’s own AWS IAM, AWS KMS, key-rotation, access-control, and auditing policies. Combined with encryption in transit, private-cloud deployment, and customer-controlled identity management, this supports ChurnShift’s principle:
Your Data. Your Models. Your Controls.
6. AWS Well-Architected Architecture and Foundational Technical Review
ChurnShift has designed a reference architecture that implements requirements across all six pillars of the AWS Well-Architected Framework:
Operational Excellence
Security
Reliability
Performance Efficiency
Cost Optimization
Sustainability
These architectural principles are incorporated into the design and deployment of the ChurnShift platform to support secure, reliable, efficient, and operationally sustainable customer environments.
ChurnShift has also successfully completed the applicable AWS Foundational Technical Review (FTR) requirements for its reviewed software offering. The FTR evaluates AWS Partner software against AWS-defined technical requirements and best practices, including areas related to security, reliability, and operational excellence.
As a result of successfully meeting the applicable AWS technical validation requirements, the reviewed ChurnShift software offering has achieved AWS Qualified Software status.
ChurnShift is also an AWS Partner Network (APN) Advanced Tier Services Partner. AWS Advanced Tier Partners are AWS Partners that have met AWS requirements associated with areas such as AWS knowledge, experience, and demonstrated customer success.
These AWS designations complement-but do not replace-ChurnShift’s own privacy, security, and customer-control commitments.
7. AI, LLM, and Model Privacy
ChurnShift recognizes that customer privacy extends beyond raw data. Models trained or fine-tuned using customer information may themselves encode valuable customer knowledge, business processes, patterns, or expertise.
For this reason, ChurnShift’s privacy architecture is designed to protect both customer data and customer-derived models. Where supported by the applicable deployment configuration:
LLM inference is performed within the customer’s private cloud architecture.
Machine-learning training and inference occur within the customer’s private cloud environment.
Customer data is not intentionally transferred to external AI providers for independent model training.
Models trained or fine-tuned using customer data or expertise remain within the customer’s private cloud environment.
Customer-specific trained or fine-tuned models are served from within the customer’s private cloud environment.
Data and model storage are protected by the applicable customer-controlled encryption architecture.
Access to models is governed by customer-controlled identity, network, encryption, and authorization policies.
This approach is a core component of ChurnShift’s commitment to:
“Your Data, Your Models, and Your Controls.”
8. Data Retention and Deletion
Because customer data resides within the customer’s private cloud environment, customers control their applicable data-retention and deletion policies.
ChurnShift does not independently retain or back up customer production data in external ChurnShift-managed systems.
Customers can apply their own organizational requirements regarding:
Data retention.
Data deletion.
Backup policies.
Model retention.
Model deletion.
Encryption-key lifecycle and rotation.
Logging and audit-record retention.
9. Compliance and Regulatory Support
ChurnShift can be configured to support customer compliance requirements and security programs, including, where applicable:
GDPR - Supporting customer control of personal data and processes such as data deletion and the right to erasure.
CCPA/CPRA - Supporting architectures in which customer data remains under customer control and is not sold by ChurnShift.
Industry-specific requirements - Supporting customer architectures and controls relevant to frameworks and regulatory environments such as HIPAA and SOC 2.
The AWS Foundational Technical Review (FTR), AWS Qualified Software designation, and AWS Partner status are technical and partner validations and should not be interpreted as independent certifications of ChurnShift’s compliance with any particular privacy law or regulatory framework.
Compliance ultimately depends on the customer’s particular deployment, configuration, use case, operational controls, and legal or regulatory obligations.
10. Changes to This Policy
ChurnShift may update this Privacy Policy from time to time to reflect changes in our technology, architecture, security practices, services, or applicable legal and regulatory requirements.
Material updates to this policy will be communicated as appropriate before they become effective.
11. AI-Assisted Drafting and Human Review Disclaimer
This Privacy Policy was prepared with the assistance of artificial intelligence (AI) technology and has been reviewed and vetted by humans prior to publication. AI was used as a drafting and language-generation tool. The information, principles, architectural descriptions, security controls, deployment characteristics, and design details presented in this document were instructed to be generated based on the existing design, architecture, implementation, and operating principles of the ChurnShift solution, including ChurnShift's commitment to: “Your Data, Your Models, and Your Controls.” The use of AI in preparing this document does not mean that the technical architecture, security controls, privacy principles, or other representations described herein were independently created or determined by an AI system. They are intended to document and communicate the existing design and principles of the ChurnShift solution based on information, technical specifications, and instructions provided by ChurnShift. Following AI-assisted drafting, the content was reviewed and vetted by humans for consistency with ChurnShift's existing solution, architecture, design principles, security practices, and intended data-handling practices. Human review is an integral part of ChurnShift's process for validating AI-assisted documentation before publication. AI assistance was used to organize, formulate, and improve the presentation of the information. Final responsibility for the content, representations, and publication of this Privacy Policy remains with ChurnShift Inc., not the AI system used to assist in its preparation. This document should not be interpreted as AI-generated legal advice. Unless expressly stated otherwise, human review and vetting of this document should also not be interpreted as independent legal, regulatory, or compliance certification or as review by legal counsel.
12. Contact Us
If you have questions or concerns regarding this Privacy Policy, ChurnShift’s security architecture, encryption practices, or ChurnShift’s data and AI model handling practices, please contact:
privacy@churnshift.com
ChurnShift Inc. “Your Data, Your Models, and Your Controls.”