Article body
Full article
Analytics often needs to sit inside a CRM, ERP system, industry SaaS product, or operations portal. A maintainable implementation gives data, metrics, permissions, and the host interface clear delivery boundaries. HENGSHI SENSE provides BI PaaS capabilities that partners can embed at several levels. With the right data preparation and permission configuration, Data Agent can serve as a conversational analytics entry point, while HENGSHI CLI provides a command interface for reviewable automation. This article explains the capabilities, implementation approach, and operating boundaries involved in bringing HENGSHI analytics into a business system.
1. Why Business Systems Need Embedded Analytics
Many business systems already hold transaction, customer, inventory, or operations data. Teams still export data, compile it offline, and rely on fixed reports to answer routine questions. Placing analytics in the business workflow lets users examine data around the current object, role, and business action without switching between systems.
Sustainable embedded analytics requires more than page integration. A working scenario needs clear data sources, metric definitions, role-based visibility, and a plan for how business users will interpret and use the results. HENGSHI provides analytics and delivery capabilities. Customers and partners remain responsible for designing and approving their industry rules, business processes, and final decisions.
2. HENGSHI Open Capabilities
2.1 Embed HENGSHI SENSE at the Right Level
HENGSHI SENSE is a BI PaaS platform for business scenarios. Partners can use iframe-based and related integration methods to embed the full platform, one module, a function within a module, a single application, or a designated folder. They do not need to expose the complete analytics interface inside the host system. For a single application, display controls for page hierarchy, sidebars, and tabs can help the embedded surface match the host page structure.
Each level suits a different delivery goal. An operations portal can embed an application designed for viewers. Industry software can embed dataset, metric-analysis, or application-authoring functions. A team that needs a shared analytics workspace can evaluate broader platform-level integration. The available embedding method, authentication design, and parameters depend on the product version and its integration documentation.
2.2 Deliver Data, Metrics, and Visual Analysis
HENGSHI provides product capabilities for building analytical content, from data connections and dataset preparation to metric management, charts, dashboards, and applications. Metrics can give teams a shared analytical vocabulary. Atomic metrics hold calculation definitions, while business metrics can add dimensions, constraints, time axes, and related context. Centralizing key definitions can help reduce repeated definitions and the effort needed to explain them within the same scenario. Business and data owners must still confirm that each definition is correct.
2.3 Data Agent and HENGSHI CLI
Data Agent provides a conversational analytics entry point that can assist with on-demand data analysis, metric creation, and dashboard generation. Its answer quality depends on prepared datasets, fields, and metrics, along with clear industry terms, analytical rules, and prompts. Relevant staff should review important conclusions and external output. AI output remains nondeterministic and should not drive business decisions or automated actions without review.
HENGSHI CLI gives people and AI Agents one command entry point for data connections, dataset access, queries and previews, dashboard and report configuration, permission operations, and selected operations tasks. Teams can organize recurring work as reusable, reviewable command workflows. Before creating resources, granting access, or publishing content, they can use --dry-run to preview the operation and then follow their approval process.
3. Define Identity and Permission Boundaries First
Embedded analytics needs identity, permissions, and data scope to remain aligned. The project team should map users in the host system to HENGSHI users, groups, or user attributes, then configure permissions for the way each application and dataset will be used. Under the applicable permission mode, HENGSHI supports row and column permissions on datasets to control which records and fields each user can access.
Permission rules must cover the datasets that users query and view. If a downstream dataset derives from or joins an upstream dataset, the team should not assume that upstream rules propagate. Test the dataset that the application uses. Before launch, test with identities from different roles and verify embedding parameters, filters, exports, and related access paths.
4. A Practical Implementation Path
4.1 Start with a Business Scenario
Choose one bounded scenario with a clear benefit, such as a store-operations dashboard, customer-operations analysis, supply-chain exception tracking, or project performance analysis. Identify the users, the questions they need to answer, the required data, and the conclusions that require human approval. A focused pilot provides stronger evidence than an early commitment to cover every business module.
4.2 Build the Data and Metric Foundation
Confirm the data connections, update method, dataset logic, and meaning of key fields. Align core metrics, dimensions, filters, and time definitions with business owners. For Data Agent scenarios, document dataset purpose, synonyms, industry terminology, and analytical rules to reduce ambiguity. Assign owners and change processes for data changes, metric revisions, and permission updates.
4.3 Select the Embedding Level and Complete Acceptance Testing
Select the embedding level for the scenario, integrate authentication and user identity, then test permissions, data consistency, interaction quality, and error handling. Acceptance testing should verify more than whether the page opens. Each role should see only authorized data, key metrics should match the agreed definitions, and the embedded interface should fit the real workflow. Measure performance in a POC that reflects the data source, data volume, query model, concurrency, network, and deployment design instead of applying one preset target.
5. Conclusion: Bring Analytics into the Workflow
HENGSHI open capabilities help software vendors, solution partners, and enterprise teams bring data analysis, metric management, visual applications, and conversational analytics into their existing systems. A durable delivery depends on a focused scenario, reliable data and metrics, verifiable permission boundaries, and careful use of AI and automation. Interface scope, deployment requirements, available features, and performance vary with the product version, configuration, and project environment. Confirm them against current documentation and POC results before production delivery.