Hybrid Prompt Agent for Conversational Maritime AIS Data Analysis
A Hybrid Prompt Agent enabling natural-language analysis of maritime AIS data by combining query classification with dynamic prompting.
2025.06 – 2025.08
·Motivation
Conventional AIS analysis requires SQL/GIS/Python expertise, while single-prompt LLM agents cannot balance factual accuracy and analytical reasoning.
·Methodology
A two-stage agent: first classify query types (fact/aggregation vs. inference/analysis), then apply tailored prompts — compact for accuracy or chain-of-thought for reasoning — executed via safe tool-calling on AIS datasets.
·Contribution
- Hybrid prompt architecture validated on a 100-item AIS QA benchmark
- Balanced improvements in accuracy, reasoning, and efficiency
