KX Dashboards: Architecture & Key Advantages¶

This page explains how KX Dashboards is built, what makes it different from conventional BI tools, and why those differences matter for real-time, high-volume analytics.

KX Dashboards is a high-performance analytics and visualization platform built natively on kdb+, the world's fastest time-series database. Most BI tools are built for periodic reporting. KX Dashboards are built for live data — millisecond query responses, genuine push-based streaming, and direct access to the full power of kdb+/q, all within a single platform.

Architecture¶

KX Dashboards communicates directly with kdb+ processes — no heavy middleware, no row-by-row SQL engines, no disk I/O bottlenecks. Data is often already resident in memory, delivering millisecond or sub-millisecond query latency.

Data stays in kdb+ native format throughout the query lifecycle, with no serialization overhead:

Traditional:  DB → API layer → JSON → frontend → chart library

KX:           kdb+/KDB-X → dashboard (direct query)

This eliminates JSON conversion, REST payload handling, and format rehydration reducing latency, CPU usage, and concurrency overhead.

Deployment architectures¶

KX Dashboards ships in two architectures that share the same UI, components, and query languages — what differs is the path between the dashboard and your data.

Dashboards Direct is the lightweight architecture. The UI talks to kdb+ processes directly over IPC/WebSocket in a single hop, with no middleware to install, configure, or operate. It is the fastest way to go from data to dashboard and is ideal for local development, prototyping, and connecting straight to existing kdb+ processes.
Dashboards on the KX Delta Platform adds an enterprise middleware tier — App Server, KX Control, and Query Router — between the UI and the data. This buys load balancing across multiple kdb+ instances, centralized session and permission management, workflow orchestration, and audit logging, at the cost of a fuller deployment.

The same platform. The same architecture. Only the deployment changes.

Dashboards DirectKX Delta Platform

Single-hop path. Queries travel from the UI straight to a kdb+ process — no App Server, KX Control, or Query Router in between. ChartGL and Datagrid can skip deserialization entirely and apply raw kdb+ data directly, lowering CPU at high volumes.

Three-layer path. The presentation layer connects through a middleware layer (App Server, KX Control, Query Router) to the data layer. The Query Router load-balances and routes requests across multiple kdb+ instances, while KX Control handles orchestration, permissions, and audit logging.

Key advantages¶

Native streaming¶

Streaming is a first-class concept in KX Dashboards, not an afterthought. Three query modes are supported out of the box:

Streaming: Push-based, event-driven updates
Polling: Configurable pull intervals.
Static: Point-in-time snapshot queries.

Visual components can update at up to ~60 FPS in real time, with no polling loops or refresh lag.

Deep kdb+/q integration¶

KX Dashboards are native to kdb+, not merely connected to it. Users can write q queries, custom aggregations, and server-side logic directly within dashboard components, with no impedance mismatch between the query language and the storage engine.

Multi-language query support¶

A single platform serves every user profile:

Language	Best For
q / qSQL	Native, fastest execution.
SQL	Broad accessibility, lower onboarding cost.
Python (via PyKX)	Data science and ML workflows.
Visual Query Builder	Non-technical users.

PyKX — Python meets kdb+¶

PyKX embeds q inside Python, allowing direct manipulation of kdb+ objects without converting to pandas DataFrames or duplicating data in memory. This enables smooth hybrid workflows:

Python ML model  +  kdb+ time-series storage  +  KX Dashboard visualization

Built for every skill level¶

A layered interface means KX Dashboards work for everyone — from non-technical analysts to experienced kdb+ developers:

Drag-and-drop dashboard builder for rapid creation without configuration overhead.
Visual Query Builder to build queries without writing q or SQL.
Full q, SQL, and Python support for advanced users who need complete control.
ViewStates for dynamic, linked query and component behavior.
Dashboard Builder (Chat Agent) — generate dashboards from a plain-language description.

Columnar & time-series optimized¶

KX Dashboards inherit kdb+'s columnar architecture, making them exceptionally well-suited for large-scale time-series analytics — fast rolling window aggregations, efficient slicing across symbols and dimensions, and OLAP-style drilldowns in real time.

High-performance rendering¶

Designed for dense, real-time visual environments — trading floors, control rooms, IoT monitoring:

Server-side data reduction before transmission.
Streaming updates that redraw only changed regions.
Many charts running simultaneously without degradation.

Simple deployment¶

KX Dashboards Direct can run locally alongside existing kdb+ processes with no multi-tier architecture required. A single install gets you from data to dashboard with minimal overhead — ideal for rapid prototyping and development.

Enterprise-grade access control¶

A robust permission model lets administrators restrict data access at multiple levels, without separate security infrastructure:

Feature	Description
Dashboard-level permissions	View and edit access controlled per dashboard.
Data-level entitlements	Row, column, and table-level filtering.
Group-based access	Aligned to team structure.
Server-side enforcement	Data is filtered before it reaches the client.
Identity integration	LDAP, Active Directory, SSO, and RBAC support.
Audit logging	For compliance and governance.

Example

A single P&L dashboard can show a trader their own book, a desk head their full desk, and a risk manager a firm-wide view — all governed by entitlements resolved at query time.

Version control friendly¶

Dashboards are stored as JSON or XML, making them behave like code artefacts:

Store in Git or any source control system.
Full version history and change tracking.
Simplified migration across development, test, and production environments.

Share dashboards including full filter and parameter state.
Export to PDF for formal reporting.
Export underlying data to Excel for further analysis.

Industry coverage¶

KX Dashboards are industry-agnostic. Wherever there is real-time data, high volume, or complex analytical workflows, KX Dashboards apply without structural changes. The same platform. The same architecture. Only the data changes.

Industry	Representative Use Cases
Financial Services	Trading, risk, market surveillance, quant research.
Capital Markets & Exchanges	Order book analytics, market microstructure.
Energy & Utilities	Smart grids, load forecasting, renewables monitoring.
Manufacturing	Process optimization, predictive maintenance.
Industrial IoT	Real-time telemetry, anomaly detection.
Aerospace & Defense	Mission data, simulation analytics.
Telecommunications	Network performance, latency tracking.
Automotive & Mobility	Fleet analytics, connected vehicles.
Healthcare & Life Sciences	Patient monitoring, operational dashboards.
Retail & E-commerce	Customer behavior, real-time inventory.
Logistics & Supply Chain	Routing optimization, operational visibility.
Environmental & Climate	Weather data, climate modeling.