The CEFAR Annual Awards aim to honor students who have engaged in a FinTech industrial project as lined up by CEFAR Academy (i.e. CEFAR Project) for attainment of academic credits and delivered outstanding performance in the project by demonstrating their ability and skills in analysing a real-life business problem and applying extensive knowledge and concepts to derive proper solutions and deliverables for the project.


Awardee: LIU Keyi 
Project Title: AI-Based Application: PFCC TalentScan and Its Commercialisation

Project Abstract:

The PFCC Resume Screening System leverages AI to enhance recruitment by automating resume screening and candidate matching. Built on Retrieval-Augmented Generation (RAG) and its advanced variant, RAG Fusion, the system combines semantic search with generative capabilities to handle complex job descriptions. Key components include a Query Classifier for intent detection, a Query Optimizer for decomposing multifaceted queries, and a Recruiter Agent for analytical insights. Using LangChain, FAISS, and HuggingFace embeddings, it ensures efficient retrieval and transparent decision-making via a user-friendly interface. This approach addresses limitations of keyword matching, improves candidate discovery, and sets a foundation for scalable, ethical AI in hiring.


Awardee: XING Yicheng
Project Title: Internal Capital Adequacy Assessment Processes for Banks

Project Abstract:

In addition to the capital required by regulation, banks should hold additional capital to effectively manage risk. This project summarizes the regulatory capital of HSBC in 2024 and calculates its additional capital buffer. By collecting financial statements, internal data and market data, the project analyzes the practices and impact of HSBC in addressing market interest rate fluctuations and risk-weighted asset (RWA) forecast errors. The project uses a 95% confidence level to calculate the 3-month rolling interest rate impact of different currency bonds, setting the FVOCI capital buffer at US $55.86 million. For RWA buffer, the project compares the difference between HSBC's internal risk model predictions and actual RWA data, and applies the quantile adjustment method to enhance the ability to resist risk in extreme market environments, determining the RWA buffer at $3.141 billion. Thus, the total capital buffer is calculated at $3.197 billion. Finally, the project summarizes the regulatory capital, regulatory requirement ratio, and capital buffer of HSBC by referring to the Basel III framework and HKMA’s banking regulatory documents.


Awardee: MA Xingyun
Project Title: Fraud Detection with LLMs: Using GraphRAG to Understand Complex Financial data

Project Abstract:
This study explores the application of GraphRAG, a knowledge graph-based retrieval-augmented generation tool, to detect shell company risks within the SCI dataset, a complex financial database. The objective was to rapidly structure and visualize financial relationships, supporting users like auditors in fraud detection. GraphRAG was employed to generate a knowledge graph focused on SCI data, enabling rapid visualization of abstract and complex relationships to facilitate quick identification of risk patterns. The effectiveness of the tool was enhanced through prompt tuning, refining its ability to capture financial entities and risks accurately. Additionally, the knowledge graph supported querying, allowing stakeholders to quickly obtain targeted information, such as shell company traits, with evidence-based insights delivered in seconds. This approach significantly outperformed traditional methods by providing accessible and actionable insights. Key contributions include rapid data visualization, improved risk detection, and efficient access to financial insights. Future work could explore scaling to larger datasets, enabling real-time updates, and enhancing the user interface to broaden its applicability in financial fraud detection.


Awardee: CEN Baihui 
Project Title: Utilize Customer Engagement Data to Provide Personalized Products/Offers


Project Abstract:
This project aims to develop a recommendation solution for the bank that sells the right things to the right person at the right moment, namely capable of targeting valuable customers and providing personalized offers at the proper moments. By deploying machine learning and artificial intelligence and digging into the massive customer engagement data, including demographic characteristics, banking transaction information, and app usage behaviors, personalized recommendations are made based on the predicted customers' purchase intention on each product.

​​​​​​​


Awardee: HU Haodong
Project Title: Development of the Local End-to-End Retrieval Augmented Generation (RAG) System Specific to the Area of Fraud and Compliance


Project Abstract:

This project aims to develop a tool for bank's employees to enhance their work efficiency, allowing them to upload PDF documents to their system and interact with it through a UI interface similar to Chat GPT, thereby retrieving information from the documents more efficiently.

​​​​​​​


Awardee:  ZHANG Junhua
Project Title: Development of an ESG-policy Oriented Portfolio Management System

Project Abstract:
This study investigates the impact of Environmental, Social, and Governance (ESG) policies on corporate behavior and stock performance in China, amid the country's sustainability challenges and economic growth. The project examines provincial ESG policies, classifies them to understand their regional influence. By reviewing policy documents and collecting stock data, the research identifies market reactions to ESG announcements as positive, neutral, or negative, indicating the financial effects.

​​​​​​​


Awardee: LIAO Zhongmin
Project Title: Digital Wallet – Business Opportunity for Commercial Banking & Use Cases


Project Abstract:
Digital wallets have reshaped how consumers manage transactions, offering commercial banks a chance to tap into this trend. This study will assess the market for digital wallets in commercial banking, identifying key applications and trends. It aims to position commercial banks strategically within the digital wallet sector by examining integration with existing services and the impact on banking operations.


Awardee: LI Tingxuan
Project Title: Financial Fraud Prevention

Project Abstract:
The incidence of financial fraud is increasing, with complex methods such as impersonation and deceptive banking messages leading to significant financial loss and undermining confidence in the banking system. Banks are countering this with more sophisticated fraud detection systems across their operations. This project aims to investigate the most widespread forms of financial fraud and identify the banking channels they commonly exploit. Students will explore the ways in which banks can apply technological advances and strategic approaches to effectively combat these threats. Students will engage in developing protective measures that ensure customer security and uphold the banking infrastructure's integrity.


Awardee:  ZHANG Jiahao
Project Title: Green Loans Based on ESG

Project Abstract:
With increasing ecological concerns, the finance sector is shifting towards sustainable lending. This project introduces a green financing system that offers lower loan rates to businesses meeting ESG (Environmental, Social, and Governance) benchmarks. Utilizing real-time environmental performance data from APIs like those from HK Electric, the system aims to financially motivate companies to pursue greener practices.


Awardee:  JIANG Xueting
Project Title: The Application of AI & Big Data in Foreign Exchange Business 

Project Abstract:
Banks are poised to embrace a transformative shift by embedding artificial intelligence (AI) and machine learning (ML) into their foreign exchange (FX) trading functions. This shift aims to minimize the current dependency on manual labor by introducing automated systems for tasks like algorithmic trading, enhancing risk management, and efficiently managing numerous small-scale FX trades. The integration of AI and ML technologies is expected to streamline banking workflows, elevate operational efficiency, and modernize the conventional approach to FX trading.



Awardee: HAN Tianyang
Project Title: Evolution of Alpha Hotpot


Awardee: ZHANG Jiawei, Jenny
Project Title:
Visionary Risk Analytics Engine

Project Abstract:
The project envisions the creation of a sophisticated risk analytics platform that explores non-financial unstructured data, such as internet behavior and social media movements, to boost the bank's forecast accuracy in risk management. This initiative aims to break away from the traditional reliance on banking data alone for risk evaluations by integrating a broad spectrum of data types to construct a thorough risk assessment tool. The projected deliverable is a working framework that identifies significant unstructured data sources and an elementary tool for risk forecasting, offering valuable insights to refine the bank's strategies for mitigating risks and maintaining operational continuity.


Awardee:  CHEN Ming
Project Title: Build a Portfolio of Automated Process

Project Abstract:
This initiative aims to arm students with expertise in intelligent automation through the combination of RPA with AI, reflecting the skillset's increasing value in today's job market. Students will develop a set of proofs of concept and demonstrations relevant to the industry to display their technical acumen. The portfolio will consist of operational automation bots that serve as verifiable demonstrations of the students' skills for potential career paths.



Awardee: LAM Ka Kit
Project Title: Comparative Study of Digital Identity and Opportunities in  Hong Kong using iAM Smart in Banking Application as an  Example


Project Abstract:
A secure digital identity system is needed as the foundation of digital transactions. This project adopts iAM Smart function and develops a website. iAM Smart log-in integration is completed. Next stage can perform an online application form submission. This project helps develop iAM Smart system into a globally recognized identity certification system for individuals and organizations, and promote securely digital transactions and contracts.


Awardee: LOUIE Wai Tak Frank
Project Title:
VIRTUAL BANKING: VISIBLE OR INVISIBLE? Embedded banking services unintentionally in our daily  life

Project Abstract:
The HKMA issued 8 licenses of virtual bank in 2018. The virtual banks will grow up in tough soil and hard to survive due to the competitions from the existing strong incumbents in Hong Kong and strong regulatory. This project is to suggest virtual bank a way to success in Kong Hong and expand it to the oversea markets.


Awardee:  MA Chin Fung Victor
Project Title: P2M Collection Model

Project Abstract:
Traditional retail stores are also expanding their business by O2O. This project researches and analyzes the consumer and merchant behavior on mobile POS collection and create a user friendly journey, designs and customer/merchant adoption strategy of the new POS solution. This will enhance the current POS solutions for online merchants to receive payments from consumers / personal customers.