Talent Summit

UOB @ SFF

Cultivating life-long learning and talent development for a future-ready workforce

AI Capability and Community Development Strategy

Introduction Video

The video discusses our efforts in fostering a community of AI practitioners. The strategy strives to equip practitioners with a variety of skillsets across the AI product development lifecycle.

Regional AI-deathon

Image
Image
Image
Image

As part of the AI capability development strategy, the Experiential Learning program is designed to accelerate participants' learning of AI and data science-related skill sets. The regional AI-deathon is an example of an experiential learning initiative which provides opportunities for participants to apply their new skills in business use cases.

Meet the Winning Teams of the 2020 Regional AI-deathon

The Protector

Sharp Rises in Scam Cases

  • Never has Singapore and UOB saw sharp rises in scam cases for the past 3 years.
  • From Jan to Jul 2020, UOB has received more than 1,000 scam cases with a total of about SGD 44 million losses
  • Expected a historic high of over 2,000 scam cases and losses of more than SGD 80 million by end 2020.
  • Impacts faced by Bank- Lower customer’s confidence in UOB, attracts complaints and bad publicity in the media, drives away businesses and faces greater scrutiny from MAS.

The Anti-Fraud Artificial Intelligence (AF.AI) system

  • Aim - Ability to learn from past customers’ transactional behaviors and stop unauthorized or fraudulent online transactions on timely before leaving their accounts.
  • How - Leverages on AI to analyse the transaction patterns and profiles of customers, augmented with fraudulent accounts reported by other customers and information received by Police.

  • Improve Productivity
  • Enhance the confidence of customer, public and MAS
  • Positive publicity for the Bank
  • Retains customers and attract new customers
  • Deters Scammers from targeting our customers
  • Timely sharing of info with Police to stop the scammers 
  • Leading Bank to combat scams effectively

Target Groups (in sequence)

  1. Elderly customers
  2. Retail customers
  3. Wholesale Banking and other the segments
  4. Subsidiaries
  5. Other banks

Scaling out (in sequence)

  1. FAST Transfers
  2. TTs done over the counter
  3. Forged Cheques

The Protectors are from the Integrated Fraud Management (IFM) Investigation Team of Group Governance, Risk & Compliance (GGRC). The team is led by Senior Vice President, Richard Soh, Head Investigation. Team Members are Vice Presidents, Leslie Tan and Rick Koh, Fraud Investigators, and Vice President, Chia Bock Sim, Data Analyst. Leslie, Rick and Richard were former law enforcement officers.

The team was supported by their mentor, Dr Simon Goo, Executive Director of Risk and Analytics Department (RAD) and his two capable staff, Vice Presidents, Amandeep Singh and Peng Cheng Lian.

The IFM Investigation Team have investigated and provided advisories to more than 2,000 scam cases in the UOB Group since being set up in July 2017.

Besides investigations, IFM actively works with the first line and law enforcement agencies to recover funds and detect and prevent potential fraud cases that can adversely impact the Bank and its customers. This tripartite partnership is part of a wider ecosystem to manage fraud risk in today’s dynamic environment. Recently, IFM received an award from the Anti-Scam Centre (ASC) of the Commercial Affairs Department (CAD) for recovering the highest scam losses among the banks between 18 June 2019 and 17 June 2020.

AI-deathon

Richard Soh
Senior Vice President - Head Investigation, Integrated Fraud Management (IFM)
Joined UOB in July 2017 after spending 16 years in law enforcement.
Richard has won innovation awards in the public sector.

AI-deathon

Leslie Tan
Vice President, Fraud Investigator, Integrated Fraud Management (IFM)
Joined UOB in December 2008 after spending 15 years in the Singapore Police Force.
Leslie received a Certificate of Appreciation - Fraud Prevention as well as the Director, CAD Public Spiritedness Award in 2020.

AI-deathon

Rick Koh
Vice President, Fraud Investigator, Integrated Fraud Management (IFM)
Joined UOB in October 2012.
Rick has more than 10 years of Commercial Crime Investigations with the Singapore Police Force.

AI-deathon

Chia Bock Sim
Vice President, Data Analyst, Integrated Fraud Management (IFM)
Joined UOB in 2018.
Bock Sim has more than 10 Years of experience in the field of Information Technology and data analysis.

AI-deathon

Amandeep Singh
Vice President, AI expert, Risk Analytics
Joined UOB six months back.
Amandeep has more than 10 years of experience in the financial industry, mostly in the modeling space, ranging from derivative pricing to machine learning.

Testimonial on AI-deathon journey

My team came up with the AF.AI idea to protect our customers against scammers. AF.AI aims to detect and stop suspicious transactions before money leaves the bank.

We took part in the AI-deathon because we wanted to get feedback from experts and management and test the effectiveness of AF.AI. We hoped to seek support from management to implement it if we won.

Working with AI experts like Simon and his team during the competition has enhanced our knowledge in AI and increased our appreciation of the fantastic things which AI can do to better detect and prevent scams. We couldn’t believe it when we built our prototype of the AF.AI. It was the first small step toward realizing our idea!

My message to future participants is to enjoy the learning process, learn from others, never stop asking questions and, most importantly, always have an enterprising mind!

Singapore Police Force’s Anti-Scam Centre (ASC)

The ASC is the “nerve centre” for investigating scam-related crimes. Its focus is to disrupt scammers’ operations and mitigate victims’ losses. Through its multi-pronged approach of Enforcement, Engagement, Engineering and Education, more than 20 external stakeholders comprising banks, Fintech companies, telecommunication companies and online marketplaces have joined the ASC in the fight against scams.

ASC is interested in this idea and wants to support it with their data on fraudulent accounts reported by local and foreign scam victims and intelligence.

The Project team will be meeting with ASC to discuss the details of the collaboration.

Crystal Ball

  • Understanding how and why funds do well is mistakenly based on easily-available data and conventional wisdom.
  • Forecasting of AUM flows is often inaccurate because it uses a bottom-up approach based on historical sales and because markets are volatile.

The goal is to accurately predict the retail fund flows in Singapore by determining and understanding its key deciding factors using quantitative & qualitative data, with the intention of implementing this application regionally across UOBAM’s distribution network.

AI-deathon

  • To know which existing funds will be popular and when should they be marketed to achieve maximum AUM flows
  • To understand how to market existing funds to maximize their AUM flows
  • To optimally design new funds to ensure continual & sizable AUM flows
  • To accurately forecast AUM in order to generate an accurate financial forecast

  • Intermediary Sales function (UOBAM Singapore & regional subsidiaries)
  • Planning & Finance function (UOBAM Singapore & regional subsidiaries)

AI-deathon

Wong Ji-Choong
Planning & Finance, UOBAM
Ji-Choong is a strategy & planning executive with more than 20 years of experience in manufacturing and finance. He is currently involved in the development and execution of the transformational growth strategy of UOB Asset Management.

AI-deathon

Shankar Panchadcharam
Business Development, UOBAM
Shankar heads the Intermediary Team in UOBAM, responsible for marketing the firm’s investment capabilities to, and developing investment solutions for the core distribution channels in the Singapore, ASEAN & North Asian markets.

AI-deathon

Yean Chun Onn
Risk Management, UOBAM
Chun Onn has been with the Risk Management team with UOBAM Singapore for over three years. His responsibilities include leading the Investment Risk & Performance analytics function and driving the adoption of innovative risk management and data analytics practices to enhance the level of risk and performance analytics across the regions.

AI-deathon

Melanie Chua
Risk Management, UOBAM
Melanie covers the Risk & Performance analytics at UOBAM. She has been working with the fixed income as well as multi-asset teams, to drive effective review and feedback derived through quantitative methods. Melanie complements the team with her interest and experience in data analytics in the risk and performance space, to research and enhance risk insights for the organization.

AI-deathon

Ryan Tsan
Data & Digitalisation, UOBAM
As UOBAM’s resident Data Scientist, Ryan is responsible for solving complex investment problem statements using AI & machine learning. Working closely with business units across the organization, he has helped developed models that enhance the business and investment processes.

Testimonial on AI-deathon journey

The idea that eventually became Project Crystal Ball was conceived a few years ago, over an informal conversation along one of UOBAM’s corridors.

At that time, while we were excited over the idea of building a unique IP that could help us to leapfrog our competitors, our progress was limited due to the expertise, tools & resources that were available then.

Through the digital transformation efforts of UOBAM & the invaluable support provided by UOB IWG & DMO, we recognized that there is an opportunity for us to finally realize our idea to its full potential.

While the journey is an exciting one, it was not without its challenges. At one point, midway through the project, the team had to restart its efforts from the beginning due to a switch in its data sources. However, each team member’s belief in the project and the overall cause, along with mutual support for each other, helped us pull us through and kept the project going.

Our advice for future participants is:

  • An idea is never too small or too late as it could become something bigger in the future
  • Believe & have conviction in your idea as this is critical in helping you to overcome many challenges
  • Surround yourself with like-minded team members that share the same conviction

GTO

UOB's Operations Center facilitates hundreds of thousands of pages in customer documents every month. In doing so, we must process millions of pages of supporting documents, for instance, invoices and bills of lading.

Processing each transaction requires a trained/skilled bank staff to review a large number of supporting documents for credit and commercial checks and other financial/compliance verifications. To process the transaction, operations staff must find relevant data (more than 40 fields) across these non-standard/unstructured documents (more than 10 types, each of more than 30 pages) and manually verify/key them into UOB systems.

Up to 25% of these documents can benefit from some kind of automated processing.

AI-deathon

  • Build a machine learning model/system that classifies and extracts key fields from documents in order to automate and improve both the accuracy and turnaround time for operations and trade teams
  • The model will also extract clauses of interest from the document and analyze them for compliance with the bank’s policies

AI-deathon

  • -30 to 50% FTE Savings due to automation
  • Faster turnaround for UOB Customers
  • Lower skill/training requirement for operators

  • Middle office BU and back office staff users will benefit from not having to do manual data entry and detailed checks on each supporting document or manual data entry
  • Start with 1 document-centric transaction/product to establish the Machine Learning pipeline and the AI-enabled transaction workflow
  • Incrementally add more products and subsidiaries to the Machine Learning pipeline and AI-enabled transaction workflow


AI-deathon

Sim Siew Meng
First Vice President, TROC/Document Trade Team
Siew Meng has worked in Trade Finance Operations for more than 20 years and is always passionate about digitalization and process improvement. Siew Meng gets extremely excited driving process changes and innovation which help to improve the productivity and well-being of her colleagues.

AI-deathon

Ben Low
Manager, TROC/Document Trade Team
Ben is an actuarial graduate but had only worked as an actuary in the insurance sector for less than a year. Ben then joined the banking sector and worked in Trade Finance for four years since Sep 2016. Ben has a passion for data analysis, risk modeling, pricing & valuation and programming in SQL, VB and JAVA.

AI-deathon

Alex Aw
Vice President, GMET/Enterprise Workflow
Alex has worked in the technology sector in the banking industry for more than ten years. Ben is passionate about using technology to simplify and solve our tedious day-to-day tasks to make our lives easier.

AI-deathon

Godfrey Gnanasounder
Lead Developer, GMET/Enterprise Workflow
Godfrey has always been inclined towards computers and technology from a tender age. He spent four years studying computer science & engineering in university and has close to ten years of experience in the field of Information Technology. Godfrey has always been passionate in learning evolving technologies and bringing it to the best use for the organization and customers.

AI-deathon

Rohit Acharya
First Vice President, GMET/Enterprise Workflow
Rohit is a hands-on manager with 20+ years of building and breaking systems. He constantly tries to match user problems with the right technology and solution, over and above the user’s immediate needs. Rohit is constantly learning new technologies and implementing them in his personal and professional life.

Testimonial on AI-deathon journey

Banks still have a lot of tedious manual processes that are prone to human error. A “digital-first, customer-first” strategy means every organization prioritizes digitization and automation of customer touch points and acquisition channels. UOB can differentiate itself by also using technology to make a quantum leap in the productivity, efficiency and accuracy of its back-office process.

At Technology and Operations, we always have ideas to use new technologies to make a difference to our customers. The AI-deathon has given us an incubator for our ideas and allowed us to take full control of our solution. The opportunity to pitch directly to UOB senior management inspired us and raised the bar for us to come up with a convincing MVP.

Our success has been a result of the great teamwork and seamless cooperation between our team members, and of course the encouragement, inspiration and guidance from our mentor.

Apart from rekindling our passion for doing new things, AI-deathon has also helped solidify the working relationship between Technology and Operations

We would encourage future participants to dive-in without being intimidated about the technology or worrying about the solution. Amazing things can be achieved when a team of like-minded individuals with complementary skills come together to define a problem and brainstorm a solution.

Loan Click

Currently LO generation is a heavily manual process, where the Operation staff needs to print out the approved credit documents, eyeball check all the limit information, terms and conditions, and then input the information manually into a LO template. The biggest pain point in the manual process is to identify wordy changes and content, and manually extract and input the LO.

Loan Click is a process application solution embedded in the Bank’s workflow to provide an integrated intelligence platform supported by five advanced technologies, Data Modeling, Machine Learning, Nature Language Processing, Image Identification and Robotic Process Automation, which aims to streamline the bank’s manual, intensive and time-consuming efforts in LO generation and validation processes.

AI-deathon
AI-deathon
AI-deathon
AI-deathon
AI-deathon

With the AI technologies, Loan Click is changing various aspects of the LO generation and validation process. It influences the operations models across the frontline, middle desk and back offices, as well as strategic business decisions and actions. Loan Click provides measurable cost savings, increases productivity and secures a competitive edge in the loan offering business.

  • RM, CFS, CAC, Audit
  • Plan for scaling-out at Y2022


AI-deathon

Laurel Yang Xiao Yan
Credit and Operations Audit Team, UOB (China) Internal Audit
Laurel is responsible for leading credit, operations and AML related reviews and recommendations. She has more than 24 years of working experience in Credit Administration & Operation, Branch Governance & Operation, and Operational Risk Management in Wholesale Banking.

AI-deathon

Melia Yin Huan Huan
UOB (China) Internal Audit
Melia is responsible for credit and operations audit. She has 18 years of working experience in the Financial and Banking industry, with a focus on Wholesale Banking business, operations and support functions.

AI-deathon

Samson Xie Chen
IT Audit Team, UOB (China) Internal Audit
Samson is responsible for leading technology-related audits, as well as onboarding artificial intelligence and data analytics solutions in support of audit transformation. He has more than 23 years of working experience in IT Governance and IT Audit in Banking.

AI-deathon

Felix Yao Xiang Zhe
IT Audit Team, UOB (China) Internal Audit
Felix is responsible for performing technology-related audits and supporting AI and data analytics solutions. He has four years of working experiences in Banking and Insurance, with a focus on delivering IT solutions to meet business needs as well as exploring new methods of data analytics and process automations.

Testimonial on AI-deathon journey

In recent years, UOB Group has been vigorously promoting staff awareness and enthusiasm about AI through holding various seminars, facilitating knowledge sharing, and providing Innovation Work Group events such as Hackathons, Innovation talks, and the AI-deathon contest.

Group Audit has rolled out a Three-Year Transformation Plan by introducing data analytics, audit automation and AI technology to transform and prepare auditors to be future-ready since 2017. Inspired by Group initiatives and country transformation, UOBC’s Internal Audit (“IA”) team has constantly uplifted its awareness, knowledge and audit techniques through learning and applying data analytics in support of a better and faster solution, cultivating a collaborative culture and raising a mindset to constantly seek for new ways to do things better.

For the 2019 Regional AI-deathon, the UOBC IA team responded with passion and submitted 7 AI-related ideas, from which Loan Click was shortlisted as one of the finalists and subsequently recognized as one of the Top Five Winners.

In past months, the Loan Click project team has received invaluable guidance and support, as well as vital suggestions and comments from country and group business partners. All this insightful advice and generous sharing has shaped the Loan Click project from a simple idea into an integrated and meaningful AI solution.

Keep passionate, think differently, stay confident and work hard.

Upskilling Businesses in AI and Data Science

Find out more

Explore other innovations

AI Capability and Community Development Strategy
AI Capability and Community Development Strategy
The strategy strives to equip practitioners with a variety of skillsets across the AI lifecycle.
More
TMRW
TMRW
TMRW empowers the digital generation with a full suite of banking solutions through a smart, mobile-only app.
More
UOB Cards
UOBAM Invest
UOBAM Invest powered by UOB Asset Management. Your expert robo-adviser gives you decades of our investment experience at your fingertips.
More
UOB Cards
UOB Cards
Leveraging on API to make payments simpler, safer and smarter.
More
UOB Jonus Chatbot
UOB Jonus Chatbot
Meet the virtual recruiter behind our intelligent engagement and recruitment experience.
More
UOB Mighty Insights
UOB Mighty Insights
Singapore’s first AI-based digital banking service that anticipates customers’ needs and provides them insights to make managing their finances simpler.
More
UOB Portfolio Advisory Tools
UOB Portfolio Advisory Tools
A brand new way to monitor, manage and optimise your wealth portfolio. Receive actionable recommendations backed by robust scenario analyses on the investments you hold.
More