The Intelligence Architecture
Designed to understand your business, your data, and the context in which decisions are made.
Agentic Platform
The Agentic Platform is the core system powering Genie. It combines conversational interfaces, reasoning systems, and product capabilities into a unified architecture that understands questions, interprets context, and takes meaningful action inside Databox.
Read moreAnalytics Platform
The Analytics Platform is Databox’s foundation. It manages integrations and deterministic data pipelines, while our proprietary Semantic Layer and Query Engine turn raw data into accurate, analysis-ready metrics. This trusted intelligence powers BI and Analytics features and provides core capabilities to the Agentic Platform.
Read moreThe real moat is the layer that sits between data and model.
Agentic Platform
The Agentic Platform is the core system powering Genie. It combines conversational interfaces, reasoning systems, and product capabilities into a unified architecture that can understand questions, interpret context, and take meaningful action inside Databox.
Together, these layers enable Genie to understand what the user is asking, determine how to respond, and execute the appropriate actions inside the platform.
Conversation Layer
The Conversation Layer is the primary interface between users and Genie. Users interact with Genie through natural language while the system continuously captures context about the user's current environment inside Databox. Key components include:
- Natural language input
- Real-time streaming responses
- Persistent session memory
Context-enabling modules help Genie understand not only the question being asked, but also what the user is currently working on.
Intelligence Layer
The Intelligence Layer interprets the customer's objective using contextual awareness, translating it into a multi-step strategy via its reasoning engine to deliver the correct answer. The agent will not only answer the question, but will understand the reasoning behind.
- Context management
- Reasoning engine
- Intent classification
- Multi-step orchestration
Capabilities
Composable building blocks that consist of personalized tools, skills, actions and state of the art LLM models which connect the data and business objectives based on the strategy from the Intelligence Layer, producing concrete results within Databox.
- Performance analysis
- Recommendation engine
- Root cause analysis
- Workspace mutations
- Smart search
Analytics Platform
The Analytics Platform is the foundational system, managing Integrations and deterministic data pipelines that validate and enrich raw data. Our proprietary Semantic Layer maps this complex data to simple business concepts, feeding the Analytical Query Engine for guaranteed-accurate, deterministic metrics. This trusted, analysis-ready intelligence powers all BI & Analytics features, providing capabilities directly to the Agentic Platform.
Metrics Intelligence DMI
Databox Metric Intelligence (DMI) is our proprietary engine designed to understand the meaning, relationships, and performance of business metrics. By combining customer data, business context, and a hybrid data platform with advanced analytical capabilities such as forecasting, anomaly detection, and causal analysis, it translates raw metrics into actionable insights that accurately predict and improve business performance.
BI & Analytics Features
The BI & Analytics features are powered by the platform's shared understanding, enabling natural, context-aware conversations without requiring users to explain their data structure. This layer provides a rich suite of analytical capabilities:
- Metrics & KPIs
- Dashboarding and Reporting
- Performance management
- Forecasts modeling
- Data analysis
- Anomaly detection
Analytical Query Engine
The Analytical Query Engine is a dedicated, deterministic system that guarantees the math is correct; it is not an LLM guessing on your data. This powerful engine executes crucial functions for reliable, analysis-ready results like:
- Metric computation
- Period comparison
- Trend analysis
- Forecasting
Semantic Layer
The Semantic Layer is the canonical source of truth for your business data, defining the domain model, entity relationships, and metric recipes. It acts as a universal translator, resolving natural-language questions into precise analytical concepts, such as mapping "how is my revenue doing" to the exact metric, dashboard, and goal.
By enforcing business logic and permissions, this layer grounds the agent's decisions and ensures every output is accurate, consistent, and safe.
- Domain enrichment
- Replication schema
- Entity-replication relationships
- Metric definitions (recipes)
- Custom provided context
- Metric equivalents
Data Platform
The Data platform is where raw data becomes analysis-ready intelligence before the LLM ever sees it, managing Replications, ETL, and warehousing. It transforms data into structured datasets with predefined data schemas. This guarantees a clean, high-context, and trusted foundation for all subsequent analysis and reasoning.
- Replications
- Reporting API
- Entity store
- Pipelines
- Warehouse
- ETL
Data Integrations
The Integrations layer delivers secure, pre-built connectors to cloud apps (HubSpot, Salesforce, Google Analytics), databases, warehouses, and APIs. It handles authentication, rate limits, schema changes, and incremental loading—ensuring reliable, uninterrupted data flow into the Data Infrastructure.
Experience Genie
An AI analyst you can trust - available 24/7, inside Databox or through your own AI tools via MCP.
