TopoPilot: Reliable Conversational Workflow Automation for Topological Data Analysis and Visualization

Abstract

Recent agentic systems demonstrate that large language models can generate scientific visualizations from natural language. However, reliability remains a major limitation: systems may execute invalid operations, introduce subtle but consequential errors, or fail to request missing information when inputs are underspecified. These issues are amplified in real-world workflows, which often exceed the complexity of standard benchmarks. Ensuring reliability in autonomous visualization pipelines therefore remains an open challenge. We present TopoPilot, a reliable and extensible agentic framework for automating complex scientific visualization workflows. TopoPilot incorporates systematic guardrails and verification mechanisms to ensure reliable operation. While we focus on topological data analysis and visualization as a primary use case, the framework is designed to generalize across visualization domains. TopoPilot adopts a reliability-centered two-agent architecture. An orchestrator agent translates user prompts into workflows composed of atomic backend actions, while a verifier agent evaluates these workflows prior to execution, enforcing structural validity and semantic consistency. This separation of interpretation and verification reduces code-generation errors and enforces correctness guarantees. A modular architecture further improves robustness by isolating components and enabling seamless integration of new descriptors and domain-specific workflows without modifying the core system. To systematically address reliability, we introduce a taxonomy of failure modes and implement targeted safeguards for each class. In evaluations simulating 1,000 multi-turn conversations across 100 prompts, including adversarial and infeasible requests, TopoPilot achieves a success rate exceeding 99%, compared to under 50% for baselines without comprehensive guardrails and checks.

Figures (10)

• Figure 1: A screenshot of the visualization window used to interact with TopoPilot. (A) a built-in chat interface is used to communicate with the LLM. (B) An interactive visualization pane displays the generated visualization. (C) A toolbar provides basic camera and visualization controls, as well as the ability to save screenshots, videos (for time-varying data), and auto-generated Python code.
• Figure 2: Python code generated to reconstruct a workflow for computing a simplified persistence diagram; variable names are shortened due to space constraints.
• Figure 3: Examples of topological descriptors supported by TopoPilot. (a) The graph of a 2D function $f$ with its critical points: maxima in red, saddles in white, and minima in blue. (b) $f$ with a saddle-maximum pair removed after persistence simplification. (c) Merge tree of $f$. (d)–(e) Contour tree of $f$ before and after persistence simplification. (f) Persistence diagram of $f$. (g) Morse--Smale complex of $f$. (h) A vector field with its critical points: source spirals in red and saddles in yellow. (i) A tensor field visualized by its eigenvector field with degenerate points: trisectors in pink and wedges in white.
• Figure 4: (a) Node tree diagram for showing the molecule dataset with local minima after persistence simplification. (b) Final visualization from (a) with the minima in blue.
• Figure 5: Computing the simplified persistence diagram with TopoPilot. (a) Volume rendering of the Hurricane Isabel dataset with an automatically generated transfer function using the Kindlmann colormap, shown alongside (b) a simplified persistence diagram.