Intual's Intelligent Architecture: The Complete Picture
Transforming enterprise data into actionable intelligence
Intual's platform transforms enterprise data into actionable intelligence through three powerful, integrated core components. These components work together to create a unified intelligence layer that empowers business decision-making across your organization.
Core Components
Meridian: The Foundation
Meridian is our foundational technology platform—the general-purpose, industry-agnostic infrastructure layer that powers every solution we build. It's designed to be the robust foundation upon which all Intual capabilities are built.
Datalake
Powers structured data collection from IoT devices, weather data, company applicants, cashflow, and anything that can be modeled in a spreadsheet.
- Universal Data Integration: Connect securely to any data source with pre-built connectors for databases, APIs, file systems, and enterprise applications
- Distributed Processing: Power complex data transformations and ELT workflows in a scalable, fault-tolerant environment
- Schema & Data Versioning: Customer data can change. Schema changes are replicated and versioned in our Datasets with Apache Iceberg's schema-versioning
Medialake
Handles raw uploads and unstructured documents like reports, presentations, documentation, pictures, videos, and audio files.
- Vector Embedding: When files are uploaded to Medialake, they're extracted and embedded in a vector store for intelligent search
- Document Processing: Browse documents, define pipelines, clean documents, extract content in an Intual-managed vector database
- Searchable Tags: Files are given user-configured tags which can be searched by Sentry
Behind the scenes, Meridian is built on industry standards like Apache Iceberg and Parquet, enabling powerful querying capabilities with Spark and MapReduce. It includes robust schema versioning to handle evolving data structures and cell-level governance for granular access control.
Meridian provides significant technical advantage through its "Build Once, Use Everywhere" principle. By investing in this foundational layer, we can launch new solutions with greater speed and quality since we don't need to recreate core infrastructure for each application.
Meridian in action
This illustrates Meridian's role as the foundation for data ingestion and processing—connecting databases, APIs, and enterprise systems into a single, governed layer that powers Datalake and Medialake.
The MVP view shows how connections are organized across sources, reflecting our principles of universal integration, distributed processing, and schema/version governance.
Fabric: The Unified Data Brain
If Meridian is the infrastructure, Fabric is the living, real-time model of your business—creating a true single source of truth.
- Entity Resolution: Intelligently merge records from different systems into unified entities (e.g., recognizing that contact_123 from Salesforce and an email from Gmail are the same person)
- Relationship Mapping: Build a comprehensive "knowledge graph" that allows traversal of complex connections between entities
- Metric Governance: Provide a single, reliable definition for every key business metric
- Object Modeling: Define business objects that reflect your real-world entities with their properties and relationships
Fabric enables users to connect multiple data sources and build custom mapping tables with selected columns from those sources. Our intuitive interface provides intelligent suggestions to help identify relationships between objects, creating a unified view that combines data from all connections.
Fabric is our core intellectual property. It transforms fragmented data into structured, queryable intelligence, allowing our solutions to answer complex, cross-functional questions that would otherwise be impossible to address with traditional systems.
Fabric visuals
Objects represent entities, their properties, and relationships—enabling entity resolution, relationship mapping, and metric governance within Fabric's unified model.
Domain and access configuration aligns with Fabric's governed model—ensuring consistent definitions and field-level controls across interfaces and use cases.
Sentry: The AI Assistant
Sentry is an AI-native agent framework that serves as an intelligent assistant and monitoring system for your data environment.
- Vigilant Monitoring: Proactively watches the Fabric for critical events or anomalies and delivers intelligent, real-time alerts
- Natural Language Interface: Parses complex user requests in plain English, queries the Fabric to find answers, and presents them clearly
- Agentic Orchestration: Autonomously selects and coordinates a sequence of actions using internal tools to fulfill a user's command
- Field-Level Access Controls: Ensures users see only the specific data they're authorized to access
- LLM Integration: Leverages specialized AI models for complex tasks with retrieval-augmented generation
Sentry makes our powerful platform accessible and intuitive. It transforms a passive analytics tool into an active, operational assistant, allowing any user—from an analyst to a CEO—to access business insights through straightforward conversation.
Sentry in practice
Sentry interprets intent, selects tools, and orchestrates a sequence of steps—bridging user questions to Fabric queries and operational actions.
Configurable models and retrieval enable domain-specific reasoning, grounded responses, and secure access aligned with object and field-level permissions.
User Interfaces
Portal: The Business User Interface
Portal provides a secure, personalized web application accessible on any device, where users can ask questions in a conversational interface and receive data-driven answers.
- Personalized Views: Customized dashboards and reports tailored to each user's role and needs
- Collaborative Intelligence: Work together on live data, eliminating the need for static reports and emails
- Conversational Interface: Ask questions in natural language and receive intelligent, context-aware responses
- Workflow Triggers: Initiate automated processes directly from the interface
Administrators can review KPIs, alerts, and system status directly in Portal—bringing operational visibility to the same place where users ask questions and take action.
Nexus Console: The Builder Environment
Nexus Console is where technical users like data analysts and solutions engineers can build powerful, automated workflows that any employee can execute instantly through Portal.
- Web IDE: Full-featured development environment with code editor, version control integration, and debugging tools
- Workflow Builder: Intuitive Canvas for creating sophisticated automated processes comprised of tasks
- Data Modeling: Visual tools for mapping and defining your business objects and relationships
- API Access: Comprehensive APIs for developers to build custom extensions and integrations
Console outputs power the experiences users see in Portal—linking objects, workflows, and agent capabilities to a conversational interface.
The Complete Intelligence Flow
- Ingest: Unify Siloed Data: Meridian connects to your enterprise systems and applications, providing a secure foundation for all data operations. The Datalake handles structured data while Medialake processes unstructured content.
- Analyze: Create Unified Intelligence: Fabric intelligently resolves entities and builds relationships, creating a unified data model that serves as your single source of truth. Objects are created through a guided creation flow that helps users connect data sources, select fields, and define relationships with minimal effort.
- Execute: Automate Critical Processes: In Nexus Console, technical users build workflows that combine AI agents and business logic to answer specific questions. Workflows can be triggered on a schedule (OnTimer) or based on changes in company data, with support for durable executions that ensure no workflows are dropped.
- Deliver: On-Demand Insights: Business users ask questions in Portal, and Sentry determines which workflow to run, delivering calculated results within seconds. This architecture enables users to get immediate answers to complex questions without needing to understand the technical details.
The era of static dashboards is over. Business happens in real-time, and teams need answers, not just data points. Intual's platform closes the loop between a question and its complex answer, empowering every employee with on-demand intelligence.
Why This Architecture Matters
- Eliminating Delays: By directly connecting business questions to automated, AI-driven answers, Intual removes the traditional delay between asking a question and getting an answer
- Technical Depth with User Simplicity: The platform provides powerful capabilities for technical users while maintaining an intuitive experience for business users
- Security and Compliance: Enterprise-grade security is built into every layer, with cell-level governance ensuring sensitive data is handled appropriately
- Scalability: The platform is designed to scale with the growth of data and increasing complexity of enterprise needs
- Multi-tenant Architecture: Our data layer provides soft-isolation per-tenant, with options for self-hosting or single-tenancy for customers with specific requirements