← Back to Technical blog

Technical article

Enterprise BI Security Governance: Connecting Authentication, Data Permissions, and Auditing with HENGSHI

Connect authentication, connection and application data permissions, tenant distribution, export controls, and audit records, and validate BI governance with real accounts.

Nov 16, 2026Technical blogHENGSHI13 min read
Agentic BIHENGSHI SENSE

Article body

Full article

When enterprise groups and software vendors open BI to subsidiaries, suppliers, or SaaS customers, sales details, customer information, and financial metrics enter more people’s workflows. A project must establish who can sign in, which data they can access, which actions they can perform, and how to check the objects and outcomes of those actions afterward. These relationships also need ongoing maintenance as organizations and customers grow.

For these delivery needs, HENGSHI SENSE can connect to existing enterprise authentication systems, configure data permissions at connection and application levels, organize customer resources through tenant spaces, and use system operation records to identify actors, objects, and outcomes. Teams can configure governance around identity integration, resource distribution, data access, and operation verification, then use actual accounts to validate access to summaries, details, and exports so analytics services follow business relationships.

Founded in 2016 for the Data+AI era, HENGSHI has pioneered a new generation of Agentic BI. Its enterprise analytics platform, HENGSHI SENSE, works with the HENGSHI JARVIS digital employee, the HENGSHI CLI execution engine, and HENGSHI BOX, a privately deployed control hub, to form an integrated service portfolio. With a decade of BI expertise and its “Powered by Hengshi” ecosystem strategy, HENGSHI empowers software vendors across industries and provides enterprise groups in manufacturing, finance, retail, government, and other sectors with integrated AI analytics for business management and intelligent decisions, built on trusted data within their private environments.

Security governance must answer four concrete questions: who is accessing the system, whose data is being used, which actions are allowed, and whether affected objects and outcomes can be traced afterward.

1. Build a usable business identity after connecting authentication

Enterprise portals and business systems usually already have account systems. HENGSHI supports authentication methods including LDAP, CAS, SAML2, OAuth2, JWT request parameters, Feishu, DingTalk, and WeCom. Enterprises can select a path for each entry point: office portals can use the corporate authentication center, while analytics pages embedded in business systems can establish a HENGSHI login session through SSO or JWT.

User mapping is another implementation priority. Administrators must identify the field that consistently represents a person and determine which fields map to roles, user groups, organizational structures, and user attributes. For example, a region code verified by the business system can become a user attribute used in data access rules. After sign-in, the platform can translate “East China regional manager” into the appropriate data scope.

Testing should include changes of account and organizational attributes to check the mapped identity and resulting data. Matching display names do not prove that accounts are identical. When one person signs out of the host system and another signs in, the embedded page must reflect the new identity. Successful authentication is the starting point for authorization.

2. Data permissions: distinguish authorization targets and execution identities

2.1 Manage data usage at the connection level

HENGSHI supports different grants for connections, directories, and tables: name visibility, structure visibility, read-only access, and full access. Data teams can first enable discovery and schema inspection, then allow data reading as business needs require. Row and column permissions can be configured in a table’s read-only access rules when finer control is needed.

Row permissions support fixed conditions, user attributes, and expressions. Cross-table authorization can also use SQL expression filters: for example, looking up a user’s regions in a regional-manager mapping table and restricting visible order records accordingly. This is easier to maintain than duplicating connections and reports for every region.

Column permissions determine which users cannot see specified fields. An enterprise can let operations staff view order trends while restricting sensitive fields they do not need. Row and column scopes must be designed separately; hiding a table column on a page does not establish data access control.

2.2 Decide whose data an application uses

Before publishing or distributing an application, choose its data permission mode. HENGSHI distinguishes three data identities: application author, dataset author, and current user. Each supports different delivery responsibilities.

Data permission modeData identity usedImplementation checks
Application authorApplication creatorThe author’s data scope and the application’s data distribution rules for recipients
Dataset authorDataset creatorEach dataset author’s access scope and the application’s row and column permission rules
Current userUser currently accessing the applicationThe visitor’s own connection and data access grants

For example, headquarters can prepare a common sales application using an author’s data and distribute regional data to users. If analysts must query according to their own connection permissions, evaluate current-user mode. Administrators should validate results against actual configuration; the ability to open an application does not prove its data scope is correct.

Configuration locations also differ. Application-author and dataset-author modes support dataset row and column permissions inside the application. Current-user mode uses the current user’s connection permissions and does not add the application’s row permissions on top. Confirm the mode before choosing where to configure rules.

2.3 Govern combined rules and derived datasets

More rules do not necessarily mean tighter restrictions. Within the corresponding authorization scope for the same user and object, multiple row permission rules form a union. If one permits Beijing and another permits Shanghai, the user can access both. Row and column permissions together constrain returned data. Administrators must inspect the combined outcome of individual grants, group grants, and inherited rules.

Permissions after data transformation also need attention. Row permissions on datasets A and B do not automatically propagate to dataset C created by joining them. To protect C, configure and validate permissions on C separately. Data model maintenance and permission maintenance should belong to the same change process.

3. Multi-tenant delivery: shared platform capabilities and customer analytics spaces

HENGSHI’s multi-tenant capabilities suit a platform serving multiple enterprise customers. The platform creates and manages tenants and shares connections, data packages, or applications with selected tenants. Tenant administrators then authorize their own members. The platform grants resources to a tenant; its administrator manages further distribution inside that tenant.

This lets the platform maintain common data and analytics assets while customers model and analyze in separate spaces. Platform resources are distributed from the platform to tenants by default. When customers need to build on platform templates, they can create independent content in their own spaces if copying or template export is permitted, avoiding direct changes to platform originals.

An implementable distribution path: maintain data and applications on the platform → authorize selected tenants → let tenant administrators configure internal users → validate data and operation scopes under actual identities.

Multi-tenancy can work with OEM customization. The former defines resource and administration boundaries; the latter handles branding, styling, and portal experience. Project teams should design both separately. Changing colors and a logo does not change data authorization.

4. From viewing to export: include data exits and auditing in delivery

Permission to view a chart does not necessarily mean an enterprise wants a user to export every detail. HENGSHI security policies support export controls, export row limits, watermarks, and HMAC signature protection for public links. System operation records can be searched by actor, time, action, object, and outcome.

Governance objectiveConfigurable HENGSHI capabilitiesAcceptance checks
Allow chart viewing while restricting data extractionExport switches, permitted export users, and row limitsTest details, aggregate results, and relevant push scenarios as actual recipients
Improve traceability after screenshots and files are forwardedPage watermarks and PDF, PPT, and PNG export watermarksCheck global settings, application settings, and actual exported files
Protect signed filter conditions in public links against changesHMAC signature protectionKeep signing keys on trusted servers and verify modified parameters fail the original signature check
Investigate a data exportSystem operation recordsSearch “export data” for the actor, exported object, and data volume

Watermarks help identify distribution sources, HMAC protects specific link parameters, and row and column permissions control data scope. Each mechanism has a distinct role. Validate their combination for the entry points in use and retain evidence of both a successful access and a restricted operation.

5. Example design: deliver the same sales analytics to multiple suppliers

The following is a hypothetical design. A retail platform plans to provide sales analytics to different suppliers. It wants common definitions for net sales, return rates, and reporting periods while letting each supplier see only its own products and continue analysis in its own space.

  1. Maintain shared definitions on the platform. The data team manages product, order, and refund relationships, confirms metric calculations, and creates a common sales analytics application.
  2. Configure tenant resources for each customer. Create supplier tenants, configure their available data scopes, and share the corresponding data packages and applications. Supplier codes should come from trusted identity and authorization configuration.
  3. Let tenant administrators allocate responsibilities. Managers receive the appropriate tenant-level summary scope; regional staff receive data access for their roles. Column permissions restrict recipients of sensitive fields.
  4. Check data exits as well. Restrict detail exports for roles that do not need them, configure watermarks, and use system operation records to verify actual actions.

Acceptance should go beyond showing a chart that looks normal. Sign in separately as supplier A, supplier B, and platform maintenance staff. Validate consistent metric definitions, data scopes, further authorization within tenants, and whether changed filters can bypass existing grants. After model changes, review permissions on new datasets.

6. Maintain the data relationships behind enterprise security

HENGSHI combines data governance with analytics delivery: enterprises can integrate existing identities, manage access at connection and application levels, distribute common analytics resources to tenants, and govern usage through export controls and operation records. These configuration relationships must evolve with organizational structures, data models, and customer delivery so analytics can be reused across more roles and customers.

HENGSHI SENSE

Resources, ecosystem, and implementation stories

Explore how teams design and ship analytics with HENGSHI.

Request a trial

Enterprise deployment, embedded delivery, and trial requests can all be handled quickly.