弹出主菜单
Main content start
Change.每天多學一點 改變．可大可小

Accounting & Finance FinTech and Financial Analytics

  1. 主页
  2. 课程
  3. Certificate for Module (Applications of GenAI and LLMs in FinTech and Business)

Certificate for Module (Applications of GenAI and LLMs in FinTech and Business)
證書(單元 : 生成式人工智能及大型語言模型於金融科技與商業的應用)

Course Code
FN174A
Application Code
2390-FN174A
Credit
6
Study mode
Part-time
Start Date
13 Jun 2026 (Sat)
Next intake(s)
Aug 2026
Duration
30 hours
Language
English
Course Fee
Course Fee: HK$10,800 per programme (* course fees are subject to change without prior notice)
Deadline on 01 Jun 2026 (Mon)
Enquiries
2867 8331 / 2867 8424
2861 0278
finedec@hkuspace.hku.hk
Apply Now
  • Add to Folders

Accept new applications for Jun 26 intake! There will be practical classes in the computer laboratory.

Apply Now

Highlights

The programme aims to provide students with contemporary knowledge of generative artificial intelligence (GenAI) and large language models (LLMs). It equips students with an essential understanding of transformers, attention mechanisms, embedding techniques, mixture of experts (MoE), fine-tuning techniques, diffusion models, speech recognition and computer vision. The programme also covers the use of computational tools and software for GenAI and LLMs, and explores their applications in financial technology (FinTech) and business contexts.

Programme Details

On completion of the programme, students should be able to 
1.    critically evaluate the technological elements of generative artificial intelligence (GenAI) and large language models (LLMs);
2.    illustrate the use of transformers, attention mechanisms, mixture of experts (MoE) and diffusion models, and the applications of embedding and fine-tuning, speech recognition and computer vision in the financial technology (FinTech) and business contexts;
3.    apply computational tools and software for GenAI and LLMs to enhance the finance and business operations; and
4.    discuss various applications of GenAI and LLMs in the FinTech and business contexts.

Application Code 2390-FN174A Apply Online Now
Apply Online Now

Days / Time
  • Saturday, 1:30pm - 7:30pm
Duration
  • 30 hours per programme
Venue
  • Hong Kong Island Learning Centre
  • Kowloon West Campus
  • Kowloon East Campus

Modules

Course Content :

(1) Foundational concepts of generative artificial intelligence (GenAI) and large language models (LLMs) for financial technology (FinTech) and business

  • Introduction to computational tools and software for GenAI and LLMs
  • Use of GenAI, ethics of artificial intelligence (AI) and data privacy
  • Prompt engineering, tokenisation, context window management, and LLM output evaluation
  • Basics of transformers, attention mechanisms and embedding techniques
    • Transformer and GenAI
    • Self-attention mechanism
    • Multi-head attention (MHA)
    • Multi-query attention (MQA) and group query attention (GQA)
    • Cross-attention
    • Multi-head latent attention (MLA)
    • Sparse attention and efficient variants
    • Feedforward neural network (FFNN) in transformers
    • Positional embedding
    • Rotary position embedding (RoPE)
    • Tokenisation and word embeddings
    • Patch embedding
  • FinTech and business applications of transformers, attention mechanisms and embedding techniques
    • Algorithmic trading signal generation
    • Document summarisation and processing
    • Customer sentiment analysis
    • Multi-faceted risk assessment
    • Real-time recommendation systems
    • Transaction sequences for fraud detection
    • Time-series forecasting of asset prices
    • Multimodal financial report analysis
    • Analysis of regulatory compliance documents
    • Semantic search of financial news for the latest market events
    • Processing scanned invoice images for automated data entry
    • Sequence modelling in transaction data

(2) Mixture of experts (MoE) and vision transformers for FinTech and business

  • Basics of MoE and vision transformers
    • Sparse MoE and vision-MoE
    • Routing algorithms
    • Vision transformer (ViT)
    • Vision-and-language pre-training (VLP)
    • Cross-layer token fusion
    • Global average pooling (GAP) in vision networks
  • FinTech and business applications of MoE and vision transformers
    • Personalised business content generation
    • Visual product catalog analysis and tagging
    • Information extraction of financial chart images
    • Automated social media content moderation
    • Multi-resolution document analysis
    • Final feature aggregation for image classification in insurance claim assessment
  • Model scaling and deployment considerations for MoE and ViT

(3) Parameter-efficient fine-tuning (PEFT) and diffusion models for FinTech and business

  • Fundamentals of low-rank adaptation, fine-tuning techniques and diffusion models
    • Low-Rank Adaptation (LoRA) and Quantised LoRA
    • Adapter layers
    • Prompt tuning and prefix tuning
    • Denoising diffusion probabilistic models (DDPMs)
    • Forward and reverse diffusion process
    • Diffusion transformer (DiT) and DiTBlock
    • Variational autoencoder (VAE)
    • Generative adversarial network (GAN)
    • Vector quantised VAE (VQ-VAE) and VQ-VAE-2
  • FinTech and business applications of low-rank adaptation, fine-tuning techniques and diffusion models
    • Adaptation of a base LLM to specialised financial jargon for question and answer (Q&A) tasks
    • Custom sentiment models for specific industry verticals
    • Finite scalar quantisation (FSQ)
    • Steering model outputs for consistent brand voice in automated report writing
    • Generation of synthetic financial time series data for stress testing
    • Synthetic data generation for training fraud detection systems
    • Anomaly detection in transaction patterns
    • Generation of realistic product images for e-commerce
    • Discrete latent representation for efficient generative modelling for market regimes
    • Modern alternative to VQ for discrete latent modelling in finance 
  • Practical PEFT workflow and governance
    • Data preparation
    • Baseline vs. PEFT evaluation

(4) Speech recognition and computer vision for FinTech and business

  • Principles of speech and audio processing, computer vision and recommendation, and related techniques for FinTech and business
    • Automatic speech recognition (ASR)
    • Connectionist temporal classification (CTC)
    • Recurrent cross-attention clustering
    • Pixel shuffle
    • Computer vision for natural language processing (NLP) augmented tasks
    • Neural recommendation systems
    • Cross-attention in multimodal fusion
    • Unsupervised learning for embeddings
    • Model quantisation and pruning for deployment
    • Retrieval-augmented generation (RAG)
  • FinTech and business applications of speech and audio processing, computer vision and recommendation, and related techniques
    • Real-time analysis and sentiment scoring
    • Speaker diarization on business conference calls for attribution
    • Resolution improvement for scanned business documents
    • Extraction of structured data from unstructured forms and receipts
    • Next-best-product recommendation
    • Personalised financial news feeds
    • Report analyser linking CEO statement to balance sheet
    • Search for historical market contexts in relation to the present day
    • Deployment of lightweight models on mobile devices for real-time expense categorisation
    • AI-powered financial research
  • Multimodal integration and real-time deployment: combining speech, vision, and recommendation outputs

Assessment method: In-class Exercise + Group Project Presentation

 

Award

Upon successful completion of the programme, students who have passed the assessments with attendance no less than 70% will be awarded within the HKU system through HKU SPACE a "Certificate for Module (Applications of Generative Artificial Intelligence and Large Language Models in Financial Technology and Business)".

Class Details

Timetable

Lecture

Date

Time

1

13 Jun 26 (Sat)

13:30 - 19:30

2

20 Jun 26 (Sat)

13:30 - 19:30

3

27 Jun 26 (Sat)

13:30 - 19:30

4

4 Jul 26 (Sat)

13:30 - 19:30

5

11 Jul 26 (Sat)

13:30 - 19:30

Teacher Information

Mr. Ray Yip

Background

Mr. Ray Yip offers a rare perspective at the intersection of Technology and Law, holding a dual academic background in Computer Science (MSc) and Law (LLB, PCLL) from HKU. His technical expertise spans Deep Learning, LLM, and Vector Search, backed by years of hands-on experience building full-stack AI systems with Python, TypeScript, and PyTorch. As an AI practitioner, Ray has extensive experience in engineering a production-ready Agentic RAG platform capable of reasoning over messy knowledge bases. Beyond text, Ray has developed Computer Vision models for medical imaging and 3D model prediction pipelines. He teaches from a "production-first" perspective, empowering students to move beyond theory and build robust and usable AI applications for the Financial and Professional sectors.

Fee

Application Fee

HK$150 (Non-refundable)

Course Fee
  • Course Fee: HK$10,800 per programme (* course fees are subject to change without prior notice)

Entry Requirements

Applicants should hold an Advanced Diploma, Higher Diploma or Associate Degree awarded by a recognised institution where the language of teaching and assessment is English. Those with a background in business, accounting, finance, economics, mathematics, statistics, science, engineering, IT or computer science would have an advantage.

Applicants with other equivalent qualifications will be considered on individual merit.

**Please upload copy of HKID and proof of degree while applying online. 

Apply

Online Application Apply Now

Application Form Download Application Form

Enrolment Method
Payment Method
1. Cash, EPS, WeChat Pay Or Alipay

Course fees can be paid by cash, EPS, WeChat Pay or Alipay at any HKU SPACE Enrolment Centres.

2. Cheque Or Bank draft

Course fees can also be paid by crossed cheque or bank draft made payable to “HKU SPACE”. Please specify the programme title(s) for application and applicant’s name. You may either:

  • bring the completed form(s), together with the appropriate course or application fees in the form of a cheque, and any required supporting documents to any of the HKU SPACE enrolment centres;
  • or mail the above documents to any of the HKU SPACE Enrolment Centres, specifying “Course Application” on the envelope. HKU SPACE will not be responsible for any loss of personal information and payment sent by mail.
3. VISA/Mastercard

Applicants may also pay the course fee by VISA or Mastercard, including the “HKU SPACE Mastercard”, at any HKU SPACE enrolment centres. Holders of the HKU SPACE Mastercard can enjoy a 10-month interest-free instalment period for courses with a tuition fee worth a minimum of HK$2,000; however, the course applicant must also be the cardholder himself/herself. For enquiries, please contact our staff at any enrolment centres.

4. Online Payment

Online application / enrolment is offered for most open admission courses (enrolled on first come, first served basis) and selected award-bearing programmes. Application fees and course fees of these programmes/courses can be settled by using "PPS by Internet" (not available via mobile phones), VISA or Mastercard. In addition to the aforesaid online payment channels, new and continuing students of award-bearing programmes with available online service, they may also pay their course fees by Online WeChat Pay, Online Alipay or Faster Payment System (FPS). Please refer to Enrolment Methods - Online Enrolment  for details.

Notes

  • If the programme/course is starting within five working days, application by post is not recommended to avoid any delays. Applicants are advised to enrol in person at HKU SPACE Enrolment Centres and avoid making cheque payment under this circumstance.

  • Fees paid are not refundable except under very exceptional circumstances (e.g. course cancellation due to insufficient enrolment), subject to the School’s discretion. In exceptional cases where a refund is approved, fees paid by cash, EPS, WeChat Pay, Alipay, cheque, FPS or PPS by Internet will be reimbursed by a cheque, and fees paid by credit card will be reimbursed to the credit card account used for payment. 

  • In addition to the published fees, there may be additional costs associated with individual programmes. Please refer to the relevant course brochures or direct any enquiries to the relevant programme team for details.
  • Fees and places on courses cannot be transferrable from one applicant to another. Once accepted onto a course, the student may not change to another course without approval from HKU SPACE. A processing fee of HK$120 will be levied on each approved transfer.
  • HKU SPACE will not be responsible for any loss of payment, receipt, or personal information sent by mail.
  • For payment certification, please submit a completed form, a sufficiently stamped and self-addressed envelope, and a crossed cheque for HK$30 per copy made payable to “HKU SPACE” to any of our enrolment centres.
More Programmes of
FinTech and Financial Analytics