What’s the problem?
The volume of scholarly literature is growing faster than any one researcher can keep up with, and making sense of it at scale is hard. Existing AI tools tend to do one narrow job well — retrieving papers or answering a question about a single document — but struggle to move fluidly from simple retrieval through to synthesising across many papers and generating new knowledge.
What’s UQ’s innovation?
ScholarAssist – led by Dr Hai Lan and Dr Tingting Wang from UQ’s School of Electrical Engineering and Computer Science – is an agentic scholarly data management system that manages and performs multi-step reasoning over large-scale bodies of scholarly work. It supports users in addressing open-ended and complex scholarly queries, going beyond simple retrieval to deep understanding, synthesis, and knowledge generation.
The system works across three levels of scholarly query: retrieving relevant papers; extracting and synthesising information across one or many papers; and, at the highest level, discovering and generating knowledge by surfacing milestone papers, mapping research trends, and proposing new directions. It does this by translating a plain-language question into an executable pipeline built from a knowledge-representation layer, a hybrid planning layer, and a unified execution layer, then running it end-to-end over the literature.
What’s the impact?
For researchers, ScholarAssist provides one place to build a knowledge base from their own papers, automatically organise it into a browsable taxonomy of problems and methods, and then interrogate it — from “summarise the methods across these papers” to “where are the gaps, and what new research could fill them?” Rather than stitching together separate tools for search, reading, and analysis, ScholarAssist aims to support the full arc of scholarly work, from finding relevant work through to generating new research ideas. A working version of the system is publicly available to try.
Interested in working with this team?
Contact the research partnerships team
Led by
School of Electrical Engineering and Computer Science, The University of Queensland
School of Electrical Engineering and Computer Science, The University of Queensland
Project Team
Professor Zhifeng Bao
Professor Zi (Helen) Huang
| AI Research Strength | Data-Centric AI Scalable & Sustainable AI |
|---|---|
| Industry Portfolio | Social and Economic Policy |
| Key Partners |
|
| Key Publication | Lan et al. (2026). AgenticScholar: Agentic Data Management with Pipeline Orchestration for Scholarly Corpora. Proceedings of the ACM on Management of Data, 4(3), 1–28. |
| Supporting Link | agenticscholar.github.io |