SEGRA: Structured Experience-Guided Graph Reasoning Agent for Gremlin Based Question Answering

2026-07-20

AI Summary

This work addresses the challenge that non-expert users struggle to formulate Gremlin queries that conform to the structural constraints of enterprise IT knowledge graphs. To bridge this gap, the authors propose SEGRA (Structured Experience-Guided Agent), which integrates graph-structural priors with historical execution experience to accurately translate natural language into Gremlin. SEGRA employs intent recognition, schema-aligned query generation, multi-example task decomposition, and execution-feedback validation. Its key innovation lies in a curriculum-guided, reusable skill library that dynamically accumulates and retrieves verified query patterns. Experimental results on an enterprise IT support benchmark demonstrate that SEGRA achieves a 7.0× performance improvement over chain-of-thought baselines. Moreover, the skill library reduces large model invocations by 20% and lowers costs by 18% while preserving answer quality.

Abstract

Enterprise IT support knowledge graphs capture rich relationships among cases, users, devices, symptoms, taxonomic categories, root causes, and historical resolutions. Yet querying them in Gremlin requires knowledge of graph schemas, traversal semantics, edge directionality, and property-graph-specific constraints, making them difficult for non-expert operators to use. We introduce SEGRA, an experience-guided agent for enterprise text-to-Gremlin question answering. SEGRA integrates intent routing, schema- and taxonomy-grounded query generation, multi-shot decomposition, execution-aware verification, and a curriculum-bootstrapped skill library that reuses verified query patterns. On an enterprise IT support benchmark, SEGRA achieves a $7.0\times$ higher mean judge score than backbone-only chain-of-thought prompting. Its skill library further reduces LLM calls by $20\%$ and dollar cost by $18\%$ relative to SEGRA without skills, while preserving answer quality. These results show that schema-grounded agent design and reusable execution experience improve both accuracy and efficiency for enterprise graph QA.

Problem

• knowledge graph
• Gremlin
• enterprise IT support
• query complexity
• non-expert usability

Innovation

• experience-guided reasoning
• schema-grounded query generation
• skill library
• text-to-Gremlin translation
• execution-aware verification