Embedded Finance Bubbles: Countering or Enabling Bias?

Traditional large banks (such as Industrial and Commercial Bank of China, JPMorgan Chase & Co., Wells Fargo, Japan Post Holdings Co. Ltd., etc.) are diligently investing in high-tech. Under very similar precepts, community banks are focusing on mergers and also investing in high-tech to counter strong disruption currents. When speaking of high-tech, large and smaller regional community banks, alike are dedicating significant percentage of operating budget and forethought towards self-disruption. According to Statista, global IT spend in banks and Securities sector is anticipated to be 547.82 billion USD by 2021. In terms of percentage, in 2020, North American banks spend up to 40% of their IT budget in new technology and 30% in Europe.

Dear Reader. "Embedded Finance Bubbles: Countering or Enabling Bias?" is a premium article. If you would like to get access to this article, please pay 1000 USD fee.

Unsupervised NLP Ushers in Sophisticated VA Era: Compliance, Trading, Issue Resolution, Information Exchange w/ Intention

Natural Language Processing (NLP) has come a long ways to be able to work with unsupervised data (self-learning) but in most industries, it is still equated to virtual assistants (‘chatbots’) that can handle basic customer queries. Most AI enabled linguistic mentions fail to distinguish between Natural Language Understanding (NLU) and Natural Language Generation (NLG) and how advanced deep learning (specifically recurrent neural networks) are ushering in a highly sophisticated era where virtual assistants go beyond answering basic customer query.

Even more amusing is how many Financial Institutions fervently market their virtual assistant as the next big breakthrough, so much so, they christen it with a human name. In this research, we move past this repeated diatribe into more sophisticated yet highly practical NLU, NLG use cases:

Dear Reader. "Unsupervised NLP Ushers in Sophisticated VA Era: Compliance, Trading, Issue Resolution, Information Exchange w/ Intention" is a premium article. If you would like to get access to this article, please pay 2500 USD fee.

PayPal Advances with AI, Biometrics and UPIs

PayPal processes 27 million transactions per day yet records an impressively low fraud rate of 0.32%, as compared to 1.32% reported by other merchants. This has been possible due to PayPal’s long standing commitment to risk management and fraud detection. PayPal is actively deploying deep learning for customer segmentation, behavioral analysis and to analyze large volume of transactions across various parameters.

Dear Reader. "PayPal Advances with AI, Biometrics and UPIs" is a premium article. If you would like to get access to this article, please pay 750 USD fee.

Deep Learning for Credit Scoring: Business Version

NeuralNet

What is Deep Learning and how does it work?

Deep Learning (DL) is inspired by ‘neural network’ or information processing pattern in human brain. Therefore, DL is not about the ability (unlike machine learning, which tends to focus on this, most of the time), but rather about the process of arriving at a conclusion.

Dear Reader. "Deep Learning for Credit Scoring: Business Version" is a premium article. If you would like to get access to this article, please pay 750 USD fee.