Anatomy of a Dashboard

Understanding how a HEAT dashboard comes to life is key to unlocking the full value of the platform. Dashboards are not just UI elements, they’re the final result of a structured, configurable data pipeline built around your data and your goals.

HEAT processes data in four distinct stages:

1. Capture – Ingest raw data from simulations or external sources.
2. Store – Place the data into a managed or external blobstore.
3. Analyze – Use runners to process and transform that data.
4. Visualise – Render insights in a dashboard tailored to your needs.

Sessions: The Foundation

All data within HEAT is grouped into sessions. A session represents a specific instance of data processing, typically corresponding to a single user session, a multi-user simulation run, event, or batch of input.

Each session is powered by a session template, which defines:

• The structure of the data flow
• What processing needs to occur
• How results are visualized

Session Templates and Nodes

A session template is a blueprint for data processing. It defines a directed graph of transformation nodes, starting from one or more input nodes representing raw data, and progressing through stages of analysis, ultimately producing outputs that drive a dashboard.

When a session is created, the template is instanced - each node in the template becomes a unique version for that session only, complete with its own configuration and output chain.

Each node passes its output to the next step. Outputs in HEAT are typed, tracked entities that link data with processing logic. This is what enables the modular, traceable nature of your dashboards.

Runners: Processing the Data

Runners are programs built using HEAT’s SDKs (available in Python, JavaScript, and C#) that fetch data, perform transformations or analysis, and emit outputs for downstream nodes.

Some runners are built-in, including:

• Preprocessing pipelines
• Timeline analyzers
• Machine learning models for cognitive metrics

You can build custom runners to support your unique data or processing needs. Runners interact exclusively through HEAT’s runtime contracts, which ensures they plug seamlessly into the pipeline.

The Dashboard Node

Every session ends (from a pipeline perspective) in a dashboard node.

The dashboard node collects insights, scores, visual elements, and metadata and passes it to HEAT’s visualization layer. This layer interprets the output and renders the interactive dashboard for the user.

You can configure the dashboard node to:

• Highlight specific KPIs
• Tailor the layout or sections
• Include or exclude certain metrics
• Attach supporting metadata

For many customers, this node can be reused across multiple session types—while others opt for custom dashboards tied to bespoke templates.

Templates and Dashboards

You can have as many session templates as needed in an environment. Each template defines a different kind of analysis workflow and can produce one or more unique dashboards depending on how it is configured.

HEAT provides a suite of standard templates as part of the HEAT Data Standard. These are plug-and-play options for simulation data that conforms to our format and provide insights out-of-the-box—no runner development required.

If your use case is specialized, you can:

• Extend a standard template
• Create a custom template
• Build custom runners
• Configure new dashboard nodes

Summary

A HEAT dashboard is not a static view—it’s the result of a flexible, robust, and extensible pipeline. It starts with raw input, moves through processing powered by runners, and ends in a visual layer shaped by a configurable dashboard node.

By defining templates and configuring nodes, you control what your data becomes—and how it’s seen.