DashChat: Interactive Authoring of Industrial Dashboard Design Prototypes through Conversation with LLM-Powered Agents

TL;DR

DashChat tackles rapid industrial dashboard prototyping by introducing a multi-agent pipeline powered by large language models that converts natural language into practical, aesthetic prototypes. Grounded in an empirically derived design space from 114 dashboards, the system orchestrates composition, assembly, and stylization through parallel agents, guided by a DAG task plan and retrieval-augmented knowledge. A TypeScript interface with Chat and History panels enables text-based interaction and iterative refinement, with data simulated for prototyping when real data are unavailable. Two user studies demonstrate that DashChat can produce high-quality prototypes quickly and with usable, flexible workflows, signaling a practical path for design ideation and client negotiation in industrial settings.

Abstract

Industrial dashboards, commonly deployed by organizations such as enterprises and governments, are increasingly crucial in data communication and decision-making support across various domains. Designing an industrial dashboard prototype is particularly challenging due to its visual complexity, which can include data visualization, layout configuration, embellishments, and animations. Additionally, in real-world industrial settings, designers often encounter numerous constraints. For instance, when companies negotiate collaborations with clients and determine design plans, they typically need to demo design prototypes and iterate on them based on mock data quickly. Such a task is very common and crucial during the ideation stage, as it not only helps save developmental costs but also avoids data-related issues such as lengthy data handover periods. However, existing authoring tools of dashboards are mostly not tailored to such prototyping needs, and motivated by these gaps, we propose DashChat, an interactive system that leverages large language models (LLMs) to generate industrial dashboard design prototypes from natural language. We collaborated closely with designers from the industry and derived the requirements based on their practical experience. First, by analyzing 114 high-quality industrial dashboards, we summarized their common design patterns and inject the identified ones into LLMs as reference. Next, we built a multi-agent pipeline powered by LLMs to understand textual requirements from users and generate practical, aesthetic prototypes. Besides, functionally distinct, parallel-operating agents are created to enable efficient generation. Then, we developed a user-friendly interface that supports text-based interaction for generating and modifying prototypes. Two user studies demonstrated that our system is both effective and efficient in supporting design prototyping.

Figures (7)

• Figure 1: Examples of Industrial Dashboards. (a) Double Eleven Product Sales Monitoring Dashboard. (b) World Population Day Global Statistics Display Dashboard.
• Figure 2: Distributions of the Domains of Industrial Dashboards.
• Figure 3: Design Space of Industrial Dashboard.
• Figure 4: (a) Distributions of View Types in Dashboard. (b) Distribution of Dashboard Layout in 2-Level Hierarchy. (c) Distribution of Analysis Tasks in Single Views.
• Figure 5: System Pipeline. (A) Textual Input to extract tasks. (B) Interactive Selections to directly assign tasks. (C) Planning tasks to determine the task workflow. (D) Integration of specific domain knowledge to augment LLM agent. (E) Invocating Agents in Parallel. (F) Transform generated output into a real prototype. (G) Select appropriate cases to enrich the knowledge base.