Data Management

2025.10

This policy outlines the requirements and controls/procedures Bioscope AI has implemented to manage the end-to-end data lifecycle, from data creation/acquisition to retention and deletion.

Additionally, this policy outlines requirements and procedures to create and maintain retrievable exact copies of electronic protected health information (ePHI) and other critical customer/business data.

Data backup is an important part of the day-to-day operations of Bioscope AI. To protect the confidentiality, integrity, and availability of ePHI, both for Bioscope AI and Bioscope AI Customers, complete backups are done daily to assure that data remains available when it is needed and in case of a disaster.

Policy Statements

Bioscope AI policy requires that

(a) Data should be classified at time of creation or acquisition according to the Bioscope AI data classification model, by labeling or tagging the data.

(b) Maintain an up-to-date inventory and data flows mapping of all critical data.

(c) All business data should be stored or replicated to a company controlled repository, including data on end-user computing systems.

(d) Data must be backed up according to its level defined in Bioscope AI data classification.

(e) Data backup must be validated for integrity.

(f) Data retention period must be defined, kept to a minimum based on business need, and comply with any and all applicable regulatory and contractual requirements. More specifically,

• Data and records belonging to Bioscope AI platform customer must be retained per Bioscope AI product terms and conditions and/or specific contractual agreements.

(g) By default, all security documentation and audit trails are kept for a minimum of seven years, unless otherwise specified by Bioscope AI data classification, specific regulations or contractual agreement.

Controls and Procedures

Data Classification Model

Bioscope AI defines the following four classifications of data:

• Critical
• Confidential
• Internal
• Public

The following sections contain definitions of each, as well as examples.

Critical

Critical data must be protected due to regulatory requirements, privacy implications, or security sensitivities.

Unauthorized disclosure of critical data may result in major disruption to business operations, significant cost, irreparable reputation damage, or legal prosecution of the company.

External disclosure of critical data is strictly prohibited without an approved process and agreement in place.

Example Critical data includes:

• PHI/ePHI
• PII
• PCI
• Production Security data, such as:
  • Production secrets, passwords, access keys, private certificates, etc.
  • Production security audit logs, events, and incident data

Confidential

Confidential and proprietary data represents company secrets and is of significant value to the company.

Unauthorized disclosure may result in disruption to business operations and loss in value.

Disclosure requires the signing of an NDA, as well as management approval.

Example Confidential data includes:

• Business plans
• Employee/HR data
• News and public announcements (pre-announcement)
• Patents (pre-filing)
• Specialized source codes
• Non-production Security data, including:
  • Non-production secrets, passwords, access keys, private certificates, etc.
  • Non-production security audit logs, events, reports, and incident data
  • Audit/compliance reports, security architecture docs, etc.

Internal

Internal data contains information used for internal operations.

Unauthorized disclosure may cause undesirable outcome to business operations.

Disclosure requires management approval. A NDA is usually required, but may be waived on a case-by-case basis.

Example Internal data includes:

• Internal documentation
• Policies and procedures
• Product roadmaps
• Most source codes

Public

Public data is information intended for public consumption. Although non-confidential, the integrity and availability of public data should be protected.

Example Public data includes:

• News and public announcements (post-announcement)
• Marketing materials
• Product documentation
• Contents posted on company website(s) and social media channel(s)

Data Handling Requirements Matrix

Requirements for data handling, such as the need for encryption and the duration of retention, are defined according to the Bioscope AI Data Classifications.

N/R = Not Required

† customer-owned data is stored for as long as they remain as a Bioscope AI customer, or as required by regulations such as HIPAA, whichever is longer. Customer may request their data to be deleted at any time; unless retention is required by law.

[^1] S3 encryption requirements are:

Encryption Algorithm: AES-256 (Advanced Encryption Standard with 256-bit keys)

• Required by NIST SP 800-175B

Data Inventory and Lifecycle Management

Bioscope AI Security team uses an automated system to query across our cloud-based infrastructure, including but is not limited to AWS, to obtain detailed records of all data repositories, including but not limited to:

• AWS S3 repositories
• AWS RDS and DynamoDB instances
• AWS EC2 volumes
• Source code repositories
• Google Workspace

The records are stored in a database system maintained by Bioscope AI security team. Records are tagged with owner/project and classification when applicable. All records are kept up to date via automation. The system is also designed to track movement of data and update/alert accordingly.

AWS S3 Object Lifecycle Management

The Bioscope AI platform will automatically adjust the storage class for certain types of data based on its usage pattern and age. This allows the Bioscope AI platform to provide competitive pricing while still allowing the customer to store large amounts of data.

AWS provides the following storage classes:

• General Purpose
• Infrequent Access
• Archive (Amazon Glacier)

S3 lifecycle policies are used to manage the storage class for certain types of data. In most cases, the Bioscope AI platform automatically adjusts the storage class but we may give customers the ability to adjust the storage class manually to meet their pricing or performance needs.

Bioscope AI performs regular full backups of all production data. We leverage S3 lifecycle policies to automatically remove old backup data. This allows older data to “age out” instead of having to explicitly delete it. S3 lifecycle policies are also used to adjust the storage class of data backups based on the age of the backup.

Other Business Data

All internal and confidential business records and documents, such as product plans, business strategies, presentations and reports, are stored outside of an employee workstation or laptop.

• Official records are stored in record management systems such as

  • Linear (tickets),
  • GitHub (source code),
  • Google Drive, Deel (HR),
  • Bill.com (expense reports), etc.

• Confidential business documents/records are be stored in encrypted form and with access control enabled on a need-to-know basis.

Transient Data Management

Data may be temporarily stored by a system for processing. For example, a storage device may be used to stage DNA raw read files prior to being uploaded to the production environment in AWS. These transient data repositories are not intended for long term storage, and data is purged immediately after use.

Bioscope AI currently does NOT use transient storage for any sensitive data.

Backup and Recovery

Customer Data

Bioscope AI stores data in a secure production account in AWS, using a combination of S3 and DynamoDB, both of which are HIPAA compliant. By default, Amazon S3 provides durable infrastructure to store important data and is designed for durability of 99.999999999% of objects.

Bioscope AI performs automatic backup of all customer and system data to protect against catastrophic loss due to unforeseen events that impact the entire system. An automated process will back up all data to a separate AWS region in the same country (e.g. US East to US West). By default, data will be backed up daily. The backups are encrypted in the same way as live production data.

Customers can also utilize the Bioscope AI Application Programming Interface (API) to extract and store their data elsewhere. Standard API usage fees will apply.

Source code

Bioscope AI stores its source in git repositories hosted by GitHub.

Source code repositories are backed up to Bioscope AI’s AWS S3 infrastructure account on a weekly basis with a common set of configuration for each repository to enforce SDLC processes.

In the event that GitHub suffers a catastrophic loss of data, source code will be restored from the backups in AWS S3.

Because AWS and GitHub can both host git repositories, we are able to leverage git’s ability to maintain a full history of all changes to our git repos via the commit log.

Business records and documents

Each data owner/creator is responsible for maintaining a backup copy of their business files local on their laptop/workstation to the appropriate location on Bioscope AI’s Google Drive. Examples of business files include, but are not limited to:

• Documents (e.g. product specs, business plans)
• Presentations
• Reports and spreadsheets
• Design files/images/diagrams
• Meeting notes/recordings
• Important records (e.g. approval notes)

Unless the local workstation/device has access to Critical data, backups of user workstations/devices are self managed by the device owner. Backups may be stored on an external hard drive or using a cloud service such as iCloud if and only if the data is both encrypted and password protected (passwords must meet Bioscope AI requirements). NOTE: No local workstations or devices are permitted to have copies of Critical data

Data Deletion Procedures

For Platform Customers

Despite not being a requirement within HIPAA, Bioscope AI understands and appreciates the importance of health data retention. Acting as a subcontractor/service provider, and at times a business associate, Bioscope AI is not directly responsible for health and medical records retention as set forth by each state. Despite this, Bioscope AI has created and implemented the following procedures to make it easier for Bioscope AI Customers to support data retention laws.

Some types of customer data may be automatically transitioned to a storage class that is appropriate for archival or infrequent usage. The guidelines for transitioning data to different storage classes is at the discretion of Bioscope AI.

Customer data is retained for as long as the account is in active status. Data enters an expired state when the account is voluntarily closed. Expired account data will be retained for 14 days. After 14 days, the project/account and related data will be removed. Customers that wish to voluntarily close their account should download their data manually or via the API prior to closing their account.

If an account is involuntarily suspended, then there is a 14 day grace period during which the account will be inaccessible but can be re-opened if the customer meets their payment obligations and resolves any terms of service violations. If a customer wishes to manually backup their data in a suspended account, then they must ensure that their account is brought back to good standing so that the API and user interface will be available for their use. After 14 days, the suspended account will be closed and the data will be permanently removed (except when required by law to retain).

For patient data as a Covered Entity

Bioscope AI is NOT a covered entity. Should we become a covered entity in the future, we would be required by law to retain healthcare records for up to 10 years beyond when service was last provided when providing healthcare services directly to patients. Any patient data that is marked for deletion will be archived for the time required by law. This archived data can be retrieved by the customer as long as it is retrieved within 10 years from date of last service.