The financial services (FS) industry is going through a period of change and disruption. Technology innovation has provided the means for financial institutions to reimagine the way in which they operate and interact with their customers, employees and the wider ecosystem. One significant area of development is the utilisation of artificial intelligence (AI) and machine learning (ML) which has the potential to positively transform the FS sector.
The future of fintech AI/ML is now
AI/ ML technology is helping fintechs and finservs to drive top line growth with smarter trading and better cross/upsell opportunities while at the same time improving the bottom line with better fraud detection and collections services. Leading financial firms are looking to capitalise on these trends and transform their businesses with an end-to-end AI strategy. AI/ ML is enabling firms to identify key insights from vast amounts of data, calculate risk, and automate routine tasks at unprecedented speed and scale utilising the power of GPU-based platforms.
AI/ML on Ubuntu
Ubuntu is the data professionals and software developers’ choice of Linux distro and is also the most popular operating system on public clouds. Ubuntu provides the platform to power fintech AI/ML – from developing AI/ML models on high-end Ubuntu workstations, to training those models on public clouds with hardware acceleration to deploying them to cloud, edge and IoT.
GPU based AI/ML computing
GPU computing is the use of a graphics processing unit (GPU) as a co-processor to accelerate CPUs for compute intensive processing. A central processing unit (CPU) usually consists of four to eight CPU cores, while the GPU consists of hundreds of smaller cores.
The GPU accelerates applications running on the CPU by offloading some of the compute-intensive and time consuming portions of the code. The rest of the application still runs on the CPU. From a user’s perspective, the application runs faster because it’s using the massively parallel processing power of the GPU to boost performance also referred to as “hybrid” computing. This massively parallel architecture is what gives the GPU its high compute performance.
AI/ ML application developers harness the performance of the parallel GPU architecture using a parallel programming model invented by NVIDIA called ‘CUDA’.
To learn more about setting up a data science environment on Ubuntu using NVIDIA GPUs, RAPIDS and NGC Containers, read the following blog
Fintech AI/ML use cases
GPU-based AI/ ML computing has many use cases in financial services like real-time fraud detection, compliance, autonomous finance, back-end process automation, supply chain finance and improving customer experience to name a few.
Artificial intelligence (AI) and automation can help bridge the gap between customer expectations and what services financial firms can offer. Forrester defines autonomous finance as algorithm-driven financial services that make decisions or take action on a customer’s behalf. Autonomous finance uses artificial intelligence (AI) and automation to deliver personalised financial advice to customers.
Robo-advisory services are algorithm based digital platforms that offer automated financial advice or investment management services and are built to calibrate a financial portfolio to the user’s goals and risk tolerance.
In times of economic uncertainty, solutions to simplify financial decisions like automated microsavings tools can help consumers increase their savings. Machine Learning algorithms not only allow customers to track their spending on a daily basis using these apps but also help them analyze this data to identify their spending patterns, followed by identifying the areas where they can save.
Combat financial crime
Financial institutions are vulnerable to a wide range of risks, including cyber fraud, money laundering, and the financing of terrorism. In order to combat these threats, financial institutions undertake know-your-customer (KYC) and anti-money laundering (AML) compliance activities to meet regulatory requirement.
AI-enabled compliance technology can reduce the cost for financial institutions to meet KYC requirements and decrease false positives generated in monitoring efforts by sifting through millions of transactions quickly to spot signs of crime, establish links, detect anomalies, and cross- check against external databases to establish identity using a diverse range of parameters.
Credit card fraud detection is one of the most successful applications of ML. Banks are equipped with monitoring systems that are trained on very large datasets of credit card transaction data and historical payments data. Classification algorithms can label events as “fraud” versus “non fraud” and fraudulent transactions can then be stopped in real time.
Many financial firms are exploring AI-based fraud prevention alternatives by building intelligent decisioning systems that derive patterns from historical shopping and spending behaviour of customers to establish a baseline which is then used to compare and score each new customer transaction.
Improved credit decisions
Lenders and credit ratings agencies routinely analyse data to establish the creditworthiness of potential borrowers. Traditional data used to generate credit scores include formal identification, bank transactions, credit history, income statements, and asset value. AI can help lenders and credit-rating institutions assess a consumer’s behaviour and verify their ability to repay a loan.
Supply chain finance
The scope and complexity of supply chains is growing fast and the relatively high cost of assessing firm creditworthiness and meeting KYC and AML requirements results in a huge trade finance gap. AI/ML has the potential to help bridge this trade finance gap.
Originators of supply chain finance now have access to a greater wealth of data about the behavior and financial health of supply chain participants. Machine learning algorithms can be applied to these alternative data- points—records of production, sales, making payments on time, performance, shipments, cancelled orders, and chargebacks to create tailored financing solutions, assess credit risk, help predict fraud and detect supply chain threats in real time and cost-effectively.
Enhanced customer experience
Advances in NLP (Natural Language Processing) mean that AI can be leveraged to provide a conversational interface with users, promising to disrupt the way customer services are delivered.
Conversational AI is enabling consumers to manage all types of financial transactions, from bill payments and money transfers to opening new accounts. By offering these self-service interactions, financial firms can free customer service agents to focus on higher-value interactions and transactions. At the heart of conversational AI are deep learning models that require significant computing power to train chatbots to communicate in the domain-specific language of financial services.
Wrapping it up
In recent years, AI/ML technology has enabled development of various innovative applications in the global financial services industry. The availability of big data, GPU hardware, parallel programming models and availability of elastic and scalable compute have been key drivers of the latest AI innovation wave.
If you are a financial institution that is embracing AI/ML to improve data-backed decisions, risk management and customer experiences, Ubuntu can be the common denominator in your AI journey from on-prem to cloud to edge.
Pic by Markus Winkler on Unsplash