Finance-led
Markets, credit, risk, regulation and disclosure define the problem. Technology serves the evidence.
Financial Technologyand AI ClusterRMIT University/Finance research/FinTech + AI Cluster
FinTech + AI translation and evidence platform
From financial uncertainty to evidence-led action.
We connect finance expertise with responsible AI, secure data systems and applied engineering—turning complex industry questions into rigorous evidence, practical pilots and deployment-ready research.
EvidenceSecuritySystems
Finance-ledResponsible by designBuilt for translation
What sets us apart
Financial problems first.
Technology with purpose.
The cluster is an industry-facing research initiative based in RMIT’s Finance discipline, connecting financial market expertise with AI, cybersecurity, data systems and software engineering.
We work with financial institutions, fintech firms, technology providers, regulators and public-sector partners—from first problem framing through pilots, industry-funded research, ARC proposals and embedded PhD projects.
Secure by design
Responsible use of AI and the security of the AI systems themselves are treated as one challenge.
Evidence to deployment
Rigorous empirical work is connected to scalable data systems and production-minded engineering.
Explore the cluster
Start with the question that brought you here.
01
Research priorities
See the questions and technical capabilities that organise our research program.
↘02Publications & evidence
Explore selected peer-reviewed work across markets, AI security and data systems.
↘03People & expertise
Find the right finance, AI or data-systems researcher for a collaboration.
↘04Data & applications
Follow the roadmap from research datasets to transparent, usable decision tools.
↘Selected publications
Research that moves between markets, models and systems.
Representative work is shown here. The full evidence library will grow through verified RMIT, USyd, ORCID, DOI and grant records.
2026Forthcoming
Markets + ML
Machine learning in corporate bonds: Evidence from China
Yiwen Fang, Xiaolu Hu, Angel Zhong, et al.
Journal of Banking & Finance↗
2026Published
Trustworthy AI
Fine-Grained Poisoning Framework Against Federated Learning
Minghui Li, Hangtao Zhang, Yanjun Zhang, Li Zeng, Chao Chen, et al.
IEEE Transactions on Dependable and Secure Computing↗
2025Published
AI Privacy
Extracting Private Training Data in Federated Learning From Clients
Jiaheng Wei, Yanjun Zhang, Leo Yu Zhang, Chao Chen, et al.
IEEE Transactions on Information Forensics and Security↗
2026Published
Graph RAG
Improving Access to Building Licensing Information in Australia: Design and Development of a Graph-Based RAG AI System
Diya Yan, Jiate Liu, Bocheng Han, Zhengyi Yang, et al.
Buildings↗
2025Published
Data Systems
Graphy’our Data: Towards End-to-End Modeling, Exploring and Generating Report from Raw Data
Longbin Lai, Changwei Luo, Yunkai Lou, Mingchen Ju, Zhengyi Yang
ACM SIGMOD Companion↗
2025Published
Markets + ML
Machine learning in the Australian equity market
Xiaolu Hu, Yiliao Song, Angel Zhong
Pacific-Basin Finance Journal↗
Selected funding and projects
Featured projects built around a real-world challenge.
Each project connects a defined problem with rigorous methods, delivery partners and a practical route to impact.
01
CSIRO Industry PhD · RMIT + CSIRO + WX Wealthverse
Risk-Aware and Trustworthy AI Agents for Adaptive Quantitative Trading
The challenge
Quantitative trading agents must adapt to changing markets without hiding risk or producing decisions that cannot be audited.
What we are doing
Multi-agent reinforcement learning, financial time-series modelling and explainable AI are combined with industry supervision in a four-year applied research program.
Expected impact
The project aims to deliver adaptive execution, real-time risk governance and interpretable trade rationales for investment, audit and compliance teams.
02
ARC Linkage · LP200200084
Defending AI-Based FinTech Systems Against Model Extraction Attacks
The challenge
Deployed financial AI can be queried by adversaries seeking to copy valuable models and expose intellectual property.
What we are doing
The team studies attack models, utility-preserving active defences and durable watermarking for both pre-deployment protection and post-deployment forensics.
Expected impact
The work is designed to reduce financial loss, protect organisational reputation and strengthen AI systems used in nationally critical services.
03
ARC Early Career Industry Fellowship · 2025
Responsible and Intelligent Data Systems for Industry Transformation
The challenge
Organisations need scalable AI-ready data infrastructure while preserving responsibility, transparency and trustworthy use.
What we are doing
Zhengyi Yang’s fellowship-backed program connects graph data management, responsible data intelligence and industry-facing system design.
Expected impact
It supports two-way university–industry mobility and turns advanced data-systems research into actionable industry outcomes and new research talent.
04
Applied AI system · Graph RAG
Graph-Based RAG for Australian Building Licensing Information
The challenge
Licensing rules are fragmented across agencies, making reliable regulatory information difficult for practitioners and the public to locate.
What we are doing
A three-layer graph-RAG system organises official documents from 17 agencies and produces context-aware, traceable responses.
Expected impact
The system improves access to regulatory information, reduces search friction and shows how responsible retrieval can support other complex policy domains.
Capabilities
One pathway from question to trustworthy system.
Machine learning
Credit, fixed income, equity markets, fraud detection and realistic model evaluation.
Text, NLP & LLMs
Narrative analytics, disclosure intelligence and responsible financial AI workflows.
AI security
Adversarial robustness, model governance, auditability and enterprise AI risk.
Graph & data systems
Knowledge graphs, graph RAG, transaction networks and scalable infrastructure.
Sustainable finance
ESG data quality, green instrument pricing, disclosure and greenwashing risk.
Research translation
Problem scoping, pilots, independent evidence and funding-ready proposals.
Our people
Cross-disciplinary by design.
Finance, AI security and responsible data systems brought together around practical financial-sector problems.
Meet all members
Core memberAssociate Professor Xiaolu Hu
Cluster Lead · Finance, RMIT University
Corporate finance, asset pricing, bond and credit markets, machine learning in finance, and sustainable finance.
View institutional profile
Core memberDr Yvonne (Yiwen) Fang
Associate Lead · Finance, RMIT University
Machine learning in financial markets, textual and narrative analysis, and sustainable finance.
Core memberAssociate Professor Chao Chen
AI & Data Analytics, RMIT University
AI security, trustworthy AI, cyber risk, large language model security, and AI-driven risk management.
View institutional profile
Core memberBenyuan Zhang
PhD Researcher · Finance, RMIT University
Responsible investment, ESG governance, regulatory compliance, and climate-related disclosures.
Affiliated memberDr Zhengyi Yang
Affiliated Member · Computer Science, University of Sydney
Scalable data systems, knowledge graphs, graph-based RAG, and responsible data-driven AI.
View institutional profileData & applications
Research people can use—not only read.
The first application roadmap turns financial reports into transparent, reproducible evidence.
Private pilot ready
Report Readability Lab
Upload a report—or locate a company filing—and generate a readability profile with document structure, language complexity, key passages and plain-language interpretation.
- Page-linked evidence
- Transparent readability metrics
- Cross-company and cross-year comparison
- Downloadable research output
ANNUAL REPORT · 148 PAGESAnalysis preview
62Readability
Key signal
Risk language becomes materially denser in the governance section.
Work with us
Bring us the problem before it becomes a specification.
Choose a pathway below or bring us an early-stage question. We will help match the problem, timeline and level of commitment to an appropriate research arrangement.
01#
Contract & commissioned research
Best suited to
For a defined business, policy or technical question that needs independent evidence or a tailored solution.
How it works
We agree the scope, deliverables and timeline, then assemble the right finance, AI and data-systems team. No student hosting is required.
02#
Industry-funded PhD
Best suited to
For a strategic challenge that needs sustained investigation and a dedicated researcher embedded around your organisation.
How it works
A full-time candidate is co-supervised by academic and industry leads over approximately 3.5–4 years, building both new knowledge and a talent pipeline.
03#
PhD internship
Best suited to
For a focused project that can test value before a larger research commitment.
How it works
A PhD researcher joins your team for 3–6 months with academic support, a specific brief and agreed outputs such as analysis, a prototype or research report.
04#
Co-funded government grants
Best suited to
For ambitious programs aligned with ARC Linkage, industry fellowships or state and Commonwealth funding priorities.
How it works
We shape the research case, partner roles, contributions and impact pathway together. Partner support can include both cash and in-kind resources.
05#
Data partnership
Best suited to
For organisations with valuable data and a need for benchmarking, modelling or new decision insight.
How it works
Data is governed through confidentiality, ethics and secure-access arrangements; the cluster returns analysis, validated methods and agreed outputs.
06#
Secondment & staff exchange
Best suited to
For teams that want research capability embedded directly in day-to-day operations—or industry practice brought into the university.
How it works
We agree a defined placement, access and knowledge-transfer plan for a cluster researcher or partner staff member, with adjunct pathways where appropriate.
07#
Executive education & training
Best suited to
For boards, leaders and technical teams building practical capability in responsible financial AI.
How it works
We co-design short courses, workshops or briefings around your context, including AI governance, LLM security, regulation and evidence-based adoption.
08#
Sponsorship & membership
Best suited to
For organisations seeking an ongoing relationship with the cluster and its research community.
How it works
Support a forum, roundtable, competition or lab—or discuss an annual relationship with priority access to events, talent and new research conversations.
Contact