Article body
Full article
The analytics value of industry software comes from a sustained understanding of business rules. Retail sell-through metrics, manufacturing delivery rates, and project-based businesses’ gross margins all require clear definitions of the measured objects, time periods, organizations, and exceptions. If those rules remain only in consultants’ experience, project code, and individual reports, each new customer requires another round of clarification, and later definition changes become difficult to track.
HENGSHI SENSE provides partners with BI PaaS capabilities spanning data modeling, metric management, visualization, and embedded integration. Partners can configure industry expertise as models, metrics, analytics applications, and task rules, then deliver them using each customer’s data connections and permissions. Industry expertise can thereby be maintained as product assets and reused in later projects.
1. Connect Industry Terminology to Executable Calculations with a Semantic Layer
A name such as “sell-through rate” alone does not tell software whether the numerator counts products with any sales or only those exceeding a sales threshold, nor whether the denominator includes every listed product or only products available for sale during the period. Partners first need to agree on the business definition and then establish relationships among products, stores, sales details, and inventory.
HENGSHI datasets and relationship models describe fields, granularity, and associations, while HQL provides reusable calculation expressions. Atomic metrics define basic aggregations; business metrics add qualifying conditions, analysis dimensions, and time axes. Partners can organize related metrics by business topic, allowing retail operations, supply-chain, or finance users to start from business language. Once the metrics are published, authorized users can use these assets in the metrics marketplace and analytics dashboards.
This separation lets industry definitions and page presentation evolve independently. When adding a channel dimension, teams first check the model and dimension relationships, then combine them in dashboards. When changing a refund rule, they return to the corresponding metric and its description and check charts that reference it. Partners can maintain calculation meaning centrally and reduce the work of copying formulas across pages.
2. Organize Industry Solutions into Four Types of Reusable Assets
| Asset | Representation in HENGSHI | Industry expertise captured by partners |
|---|---|---|
| Data models | Data packages, datasets, relationships, and parameters | Business objects, detail granularity, field mappings, and associations |
| Metric systems | Atomic metrics, business metrics, and business topics | Formulas, qualifying conditions, time axes, dimensions, and terminology |
| Analytics applications | Chart libraries, dashboards, applications, and templates | Business pages, analysis paths, presentation, and interaction |
| Rules for Agent use | Dataset knowledge, analysis examples, and supporting Skills | Common questions, synonyms, implicit rules, and repeated implementation steps |
Data packages can contain datasets, models, parameters, and chart libraries, with support for copying, replacing data connections, and importing or exporting templates. Partners can build a standard analytics framework first, then complete field mapping, connection adaptation, and customer-specific metric configuration. Application publishing, the application marketplace, and templates provide ways to distribute and use analytics results. The delivery method should match the partner’s product form and deployment plan.
Reuse also includes maintenance practices. Standard industry definitions, customer-specific extensions, and page configurations should be recorded separately, with their dependencies checked during updates. Partners can then keep upgrading the common solution while preserving business differences already confirmed by customers.
3. Deliver Analytics into Customers’ Existing Work Pages
Industry assets see continued use when they become part of everyday business. HENGSHI offers integration paths including iframe, JavaScript SDK, and APIs. Partners can choose how to embed charts, dashboards, applications, or related modules according to the required integration depth and connect user identities through approaches such as single sign-on and JWT. OEM and page customization can support the partner’s brand and product experience.
In store-management software, for example, the store homepage can show sales and inventory overviews, a product detail page can show that product’s sell-through performance, and a regional business page can summarize differences among stores. The host product passes the current store, product, organization, and time range, so the analytics area presents data around the same business object. Users can review relevant analysis while handling store operations, and partners can make analytics part of the product workflow.
Multitenant delivery requires configuring identity and data scope together. HENGSHI provides management of tenants, users, resource authorization, and row- and column-level data permissions. When reusing templates, partners still need to check each customer’s data connections, user mappings, and visible scope. The standard solution can be reused, while customer data and permissions are managed within their respective environments.
4. Make Industry Knowledge Useful to Both People and Agents
The same models and metrics can support analysts building dashboards and provide business context for Data Agent. Adding purposes, business terms, synonyms, and implicit filter rules to dataset knowledge can help the Agent locate the right data objects. When a user asks “Which stores have declining sell-through rates?”, the system has a basis for organizing a query around defined metrics, store dimensions, and time ranges rather than guessing from an ambiguous name.
For implementation automation, HENGSHI CLI’s hbi commands provide programmable operations for resources including data, models, metrics, and dashboards, with structured output and preview capabilities for some changes. Supporting Skills can guide external Agents through BI engineering steps by resource domain. Partners can turn recurring connection checks, metric validation, application configuration, and delivery checks into task rules, reducing repetitive mechanical work in projects.
Reuse occurs across data assets, analytics applications, and engineering methods. Industry metrics define calculation meaning, application templates define pages and analysis paths, and Skills guide Agents in using tools. Together, they turn consulting experience into objects that later projects can read, execute, and check.
5. An Example of Productizing a Retail Software Solution
Suppose a retail software vendor delivers store-performance analytics to several brands. The team first agrees on an illustrative “sell-through rate”: the number of SKUs with valid sales during a period divided by the number of available-for-sale SKUs included in the same period. It then specifies whether the calculation is at store, product, or regional level and how returns and product listing changes are handled. This definition illustrates a modeling method; it is not a universal industry standard.
Implementation staff map fields from each customer’s POS and product systems to the standard model, create metrics for SKUs with sales, available-for-sale SKUs, and sell-through rate, and build store overview and category-analysis templates. When onboarding a new customer, the team reuses the metric and page framework, focusing on connections, codes, definition differences, and permissions. Regional managers then check store results in their own business pages and validate the metrics through data questions using the same scope.
After finding a decline in sell-through rate, a manager can break it down by category and store, inspect the corresponding product details, and return to the business system to decide on transfers or replenishment. The partner incorporates common improvements into the next model and template version, retains customer-specific rules as extension settings, and retests using the same sample set. Industry expertise thus supports both new project delivery and ongoing use by existing customers.
HENGSHI supplies the technical foundation for data, semantics, visualization, embedding, and Agent tools, while partners encode industry understanding into those assets. Together, they can move analytics solutions from one-off delivery toward maintainable, upgradable industry products.