Embedded Analytics Architecture: Components & Best Practices

TL;DR: Embedded analytics architecture is the framework that enables analytics to run securely and efficiently inside another application. Organizations must plan tenant isolation, metric governance, data access patterns, embedding methods, and security controls before launching customer-facing analytics. Bold BI® supports embedded analytics through two embedding methods (iframe and the Embed SDK), server-generated embed tokens, row-level security, and three deployment editions (Cloud, Managed Private Cloud, and On-Premises).

What is embedded analytics architecture?

Embedded analytics architecture is the framework that enables dashboards and AI-powered analytics to operate directly inside another application. It connects data sources, processes analytics requests, enforces security policies, and delivers insights within the workflows users already use. The architecture determines how analytics scales, how customer data is protected, and how metrics remain consistent across users.

In Bold BI, embedded analytics can be delivered through two embedding methods: iframe embedding and Embed SDK embedding. This allows teams to balance implementation speed with customization requirements. Bold BI is Syncfusion®’s business intelligence and embedded analytics platform, a separate product from Bold Reports®, BoldDesk® and BoldSign®.

How embedded analytics architecture works

Embedded analytics architecture works by moving data from source systems through processing and security layers before delivering insights inside an application. Data is collected from operational systems and transformed for analysis. It is then secured through authentication and authorization controls, processed by an analytics engine, and displayed through dashboards or AI-driven interfaces.

A typical workflow order includes:

1. Data collection from connected sources.
2. Data preparation and transformation.
3. Security and access validation.
4. Analytics processing.
5. Dashboard or AI response generation.
6. Delivery within the host application.

With Bold BI, embedded dashboards are commonly secured through back-end-generated embed tokens and row-level security policies, helping organizations enforce tenant-aware access controls within customer-facing applications. In Bold BI, step 3 happens at query time. The embed token carries JWT claims, and row-level security scoped to the tenant, organization or role is applied before any data is returned, rather than filtered afterward.

How a Bold BI embedded dashboard loads:

1. The application UI asks your application server for authorization.
2. Your server, which holds the embed secret and any filters or parameters, requests an embed token from the Bold BI server.
3. The Bold BI server validates the request and generates the embed token.
4. Your server returns the embed token to the application UI.
5. The Bold BI JavaScript SDK in the browser sends the token to the Bold BI server.
6. The Bold BI server returns the dashboard data for the SDK to render.

The embed secret stays on your server and never reaches front-end code.

Core components of an embedded analytics architecture

The core components of an embedded analytics architecture include data sources, the data integration layer, the semantic layer, the analytics engine, the AI and agent layer, the visualization layer, the security layer, and embedding methods. Together, these components deliver secure, scalable, and governed analytics experiences inside applications.

Data sources in embedded analytics architecture

Data sources provide the information that powers embedded analytics. These sources can include databases, SaaS applications, APIs, cloud storage platforms, and enterprise data warehouses. Bold BI has 145 documented data connectors across 23 categories, plus a Generic Web API connector for sources not on the list.

A SaaS company may combine operational data from PostgreSQL, customer information from Salesforce, payment data from Stripe, and marketing metrics from Google Analytics. As data volumes grow, maintaining reliable and performant connectivity becomes increasingly important.

Data integration layer for analytics workloads

The data integration layer prepares data for analysis. It transforms, cleanses, aggregates, and refreshes information before users access it through dashboards. Many organizations implement data warehouses, staging environments, and caching strategies to improve dashboard performance while reducing load on operational systems.

As analytics adoption grows, this layer often becomes one of the most important areas for optimization. In Bold BI, Bold Data Hub pipelines handle connectivity, transformation, scheduled automation, aggregation and loading to a destination data store. And each pipeline automatically creates a Bold BI data source that stays live to that destination.

Semantic layer for metric governance

A semantic layer defines business metrics once so dashboards use consistent business definitions. Without a centralized approach, different teams may calculate monthly recurring revenue, churn, retention, or revenue differently, resulting in conflicting numbers across dashboards.

Centralized metric definitions improve governance, consistency, and trust in analytics. Bold BI supports reusable metric definitions through data source expressions, allowing calculations to be shared across dashboards built on the same data source.

Analytics engine for query processing

The analytics engine processes user requests and converts raw data into meaningful insights. It executes queries, applies filters, performs calculations, and determines which data users are authorized to access.

The efficiency of this layer directly influences dashboard load speed, scalability and overall user experience, and the effect grows as adoption grows. Bold BI supports both live and extract data modes, allowing organizations to balance real-time data access requirements with dashboard performance.

AI and agent layer for natural language analytics

The AI and agent layer enables users to interact with analytics using natural language. AI-powered experiences can explain trends, identify anomalies, summarize results, and answer business questions.

As organizations adopt AI-powered analytics, architectural decisions become increasingly important. AI systems must respect governance policies, security controls, and tenant boundaries while providing accurate responses. Bold BI has three AI surfaces, AI Copilot, AI Agent and AI Assistant, which share one AI engine and security model. And they operate within the same authentication, authorization and row-level security controls used by embedded dashboards.

When the Bold BI AI Assistant is embedded with the Embed SDK, its settings control what each user sees. You can hide the logo or beta tag, limit which dashboards and data sources it suggests. And you can lower the number of suggested questions from the default of 6. The assistant follows each user’s permissions.

Visualization layer for embedded dashboards

The visualization layer presents analytics to end users through dashboards, charts, KPI cards, maps, and data grids. This is the layer users interact with when exploring performance metrics and business trends. Effective visualizations should feel like a natural extension of the host application rather than a separate product. Bold BI dashboards can be white-labeled with your own branding and use responsive dashboard layouts. With the Embed SDK, they render inside your application’s page rather than in an iframe.

Bold BI dashboards can be embedded through the Embed SDK in React, Angular, Vue, ASP.NET Core, and Blazor applications. It also enforces row-level security at query time through JWT claims rather than filtering results afterward. And it can apply it at up to three layers: isolation code, custom attributes and user-based filters. In self-hosted deployments, each tenant can use its own identity provider, such as Azure AD, Okta or Auth0.

Embedding methods: iframe vs. Embed SDK

Bold BI offers two embedding methods: iframe embedding, which needs only an embed URL, and the Embed SDK, which renders dashboards inside your application’s own page and gives full API control over layout, filters and events. The right embedding method depends on how much customization, interaction, and application integration your analytics experience requires.

Many organizations begin with iframe embedding to validate demand and later move to SDK-based embedding when analytics becomes a core product feature. Bold BI supports both iframe embedding and the Embed SDK, allowing teams to start with simpler implementations and expand customization over time.

For example, a typical Bold BI embedding implementation uses the Embed SDK and an embed token generated on your back end to securely render a dashboard within an application:

var boldbiEmbedInstance = BoldBI.create({
    serverUrl: "<Bold BI Server URL>",
    dashboardId: "<Dashboard Id>",
    embedContainerId: "dashboard_container_id",
    embedToken: "<Embed token generated from backend server>"
});
boldbiEmbedInstance.loadDashboard();

Beyond dashboards, Bold BI exposes REST APIs and two MCP offerings: a Documentation MCP server, and the Bold BI MCP Tool, which gives MCP-compatible AI agents 53 tools in 11 categories for discovering, inspecting and exporting dashboards and data sources. Calls run with the permissions of the user whose API key is used.

Embedded analytics architecture for SaaS applications

Embedded analytics for SaaS applications must support tenant isolation, scalability, and security across multiple customer environments.

The right model depends on compliance requirements, customer expectations, and long-term scalability goals. Bold BI supports both single-site and multisite architectures for multitenant analytics deployments. In Managed Private Cloud and On-Premises deployments, a single Bold BI instance can serve multiple isolated tenants, each with its own database, branding, and delegated administration.

Common embedded analytics architecture patterns

Organizations typically choose architecture patterns based on data freshness, scalability, and performance requirements.

Many organizations combine multiple patterns to balance performance and freshness requirements. In Bold BI, direct query maps to Live mode and cached analytics maps to Extract mode. For extracts, refreshes can be scheduled hourly, daily, weekly or monthly (Amazon Redshift is the documented example). And newer direct connectors such as Azure Databricks support Live mode. This allows organizations to choose between real-time data access and performance-optimized analytics experiences.

Embedded analytics deployment options

Bold BI provides three deployment options: Cloud, Managed Private Cloud, and On-Premises editions. These deployment options determine where analytics runs, how infrastructure is managed, and where data resides. The On-Premises edition supports air-gapped deployments for organizations with strict governance, compliance, or data residency requirements.

• Cloud: Fully managed, with simplified administration and rapid scalability. Best for speed and operational efficiency.
• Managed Private Cloud: A dedicated cloud environment, operated for you, with more control over security, governance, and compliance.
• On-Premises: Complete control over infrastructure, data, and security policies, including air-gapped and offline operation. Best for strict governance or data residency requirements.

Deployment decisions should be based on compliance requirements, operational capabilities, and data residency policies. For a detailed comparison, see our BI deployment model guide.

Build vs. buy: Choosing an embedded analytics strategy

The build-versus-buy decision depends on available engineering resources, customization requirements, and time-to-market goals.

For many organizations, the decision comes down to resource allocation. Building means developing and maintaining dashboards, security controls, multitenancy, APIs, and supporting infrastructure. A platform lets your team spend that effort on core product innovation. Bold BI includes embedding and security controls in all three editions, with multitenancy in Managed Private Cloud and On-Premises. It has published per-user and per-application prices, with enterprise licenses quoted on request, and no separate setup fee. For a deeper look, read Choosing Embedded BI: Build or Buy? and Build vs. Buy Analytics: Which Is Better for Your Business?

Embedded analytics architecture best practices

Many embedded analytics problems come from decisions made, or skipped, early on. Use these best practices as a checklist before your first customer-facing release.

Design for multitenancy from day one

Tenant isolation becomes increasingly difficult to implement after customers are already onboarded. Select an isolation strategy early and build around it. In Bold BI, one instance can host multiple isolated tenants, each with its own database and branding, in Managed Private Cloud and On-Premises deployments.

Define metrics once in a semantic layer

Consistent metric definitions improve trust in analytics and prevent conflicting dashboards. Business-critical calculations should be defined centrally and reused. Bold BI supports reusable calculations through data source expressions.

Generate embed tokens on the server

Authorization logic should never be exposed to the browser. Server-side token generation helps reduce security risks and simplifies governance. With Bold BI, your back end generates the embed token and the embed secret never reaches front-end code. So tenant-aware filtering and access rules stay hidden from end users.

Test with two tenants before every release

Broken access control issues often appear only when multiple tenant environments are tested side by side. Before every release, validate analytics access, exports, and AI-powered interactions across separate customer accounts to confirm tenant boundaries remain intact.

Match your data access pattern to freshness needs

Not every dashboard requires real-time data. Evaluate whether operational dashboards need direct-query access or whether cached approaches can deliver better performance without affecting decision-making. Organizations using Bold BI can choose between Live and Extract data modes based on freshness and performance requirements.

Monitor query load as adoption grows

Analytics workloads often increase rapidly as adoption expands. Monitor response times, usage patterns, and expensive queries before performance issues become visible to end users. Query monitoring often identifies when caching, warehouse-based analytics, or infrastructure scaling becomes necessary.

Scope AI access to governed data

AI systems should inherit the same governance and security controls applied to dashboards. Restrict AI access to approved data sources and ensure responses respect tenant boundaries and access permissions. In Bold BI, the AI Agent enforces row-level security before retrieving any data. And the embedded AI Assistant’s dashboards and data sources settings limit what it suggests. While each user can query only the data they have permission to access.

Embedded analytics architecture example: A SaaS scheduling platform

Consider a hypothetical SaaS company that provides appointment-scheduling software to healthcare clinics. To reduce manual analytics requests, the company embeds a no-show dashboard directly within its application.

Using the Bold BI Embed SDK and a back-end-generated embed token, each clinic can securely access only its own analytics data. As adoption grows, the company expands the experience with revenue and staff-performance dashboards and later introduces AI-powered analytics.

What changes: Clinic managers gain faster access to insights, while support teams spend less time generating manual dashboards.

How Bold BI supports embedded analytics architecture

Bold BI®, developed by Syncfusion®, is a business intelligence and embedded analytics platform that helps organizations build secure, scalable, and customizable embedded analytics experiences. It supports embedded analytics architecture by:

• Offering two embedding methods: iframe embedding and the Embed SDK.
• Allowing you to embed analytics into React, Angular, VueJS with a Go backend sample and Node.js, ASP.NET Core, and Blazor applications.
• Deploying in any of three editions: Cloud, Managed Private Cloud, and On-Premises.
• Allowing multitenant analytics, with isolated tenants on a single instance in Managed Private Cloud and On-Premises.
• Providing row-level security, enforced at query time, and server-side filtering capabilities.
• Offering pricing per user, per application or as an enterprise license, with published prices for per-user and per-application plans.

Together, these capabilities help organizations deliver secure, scalable, and multitenant analytics experiences within their applications while maintaining control over deployment, security, and user access.

Ready to put this embedded analytics architecture into practice? Start a free trial or book a personalized demo to see how Bold BI handles security, multitenancy, and deployment in your environment.

Frequently asked questions

1. 1.

   What are the core components of embedded analytics architecture?

   Embedded analytics architecture typically includes data sources, data integration, analytics processing, AI capabilities, visualizations, security, and embedding methods. Together, these components deliver secure and scalable analytics within applications.

2. 2.

   Why is embedded analytics important for SaaS applications?

   Embedded analytics lets users access dashboards and insights directly within a SaaS application, eliminating the need to switch tools. This improves user experience and makes analytics a seamless part of the product.

3. 3.

   How does AI enhance embedded analytics?

   AI enables users to explore data using natural language instead of building reports manually. In Bold BI, AI features help users get insights faster while respecting existing security and access permissions.

4. 4.

   Does Bold BI charge per user for embedded analytics?

   Not necessarily. Bold BI offers per-user and per-application plans with published prices, and enterprise licenses quoted on request, so you can choose the model that fits how your customer base grows.

5. 5.

   Which deployment editions support embedded analytics in Bold BI?

   Embedded analytics is supported across all Bold BI editions: Cloud, Managed Private Cloud, and On-Premises. Organizations can choose the option that best fits their infrastructure and data residency requirements.

6. 6.

   Which application frameworks can embed Bold BI dashboards?

   Bold BI supports embedding in popular frameworks, including React, Angular, Vue, .NET, and Blazor. SDKs and sample applications are available to simplify integration.

7. 7.

   Does Bold BI support multitenant embedded analytics?

   Yes, in Managed Private Cloud and On-Premises deployments. Bold BI supports multi-tenant embedded analytics, allowing organizations to manage multiple tenants with isolated data, branding, and administration from a single instance. Cloud Edition does not support multi-tenancy.

8. 8.

   What are the current limitations of the Bold BI AI Assistant?

   The AI Assistant supports common aggregations such as COUNT, SUM, MIN, MAX, and AVG. However, it does not currently support custom expressions, dashboard editing, or querying multiple data sources within a single conversation.