Applied AI in Finance

"Silicon Valley is coming." - Jamie Dimon, CEO of J.P. Morgan

How does AI impact financial services?

AI can impact financial services in a multitude of ways - and is already impacting the sector. Artificial intelligence is changing the traditional financial services from banking and insurance to analysis and asset management.

Overall, the impact of AI on financial services can be divided into AI-based financial offerings previously offered by traditional providers, and AI-based sector agnostic solutions, or so-called horizontal solutions impacting generic processes in for example sales, retention, call center and IT operations. The latter is not exclusively developed for application within Financial Services but is merely an expression of a series of widely applicable services with multiple sector perspective. An example of an AI-based financial solutions offering is credit scoring whereas platforms such as Ayasdi goes across sectors. Further details are provided below.

In the video above the founder and CEO of GupShup talks about the new customer interface based bots that are quickly changing personal finance.

We define different use cases within the categories of Credit Scoring & Direct Lending, Assistants & Personal Finance, Quantitative and Asset Management, Insurance, Market Research & Sentiment Analysis, Debt Collection, Business Finance & Expense Reporting, General Purpose & Predictive Analytics and Regulatory, Compliance & Fraud Detection, to describe the utilization of AI within Financial Services offerings and horizontal solutions.

black_voice_control-50.svg
Assistants & Personal Finance
The use of assistants, and the so-called "bots", are widely applied cross sector to develop new customer experiences, therefore, making it a key theme within horizontal AI-based solutions. Where the company Kasisto is an example of an intelligent product offering within personal finance management, other companies such as GupShup are enabling existing companies to delegate specific service offerings to assistants utilizing voice or messaging to offer new customer experiences or entire solutions. Please refer to our video interview with CEO and Founder of GupShup above.
black_taxes-31.svg
Credit Scoring & Direct Lending

Applying AI within credit scoring and direct lending have manifested itself with new ways of performing more accurate credit scoring but also widened the use of new data sources to expand offerings to new customer segments previously considered out of scope. By utilizing AI, companies such as Avant, Vouch and Affirm have sought to expand the traditional credit scoring, with the latter focusing especially on expanding its offerings to customers who would otherwise not be eligible for point-of-sale consumer loans.

As part of the Applied AI Project, we at Innovation Centre Denmark Silicon Valley, interviewed Ryan Metcalf, Chief of Staff & Director of International Markets at Affirm, about AI. Their mission is simple: deliver honest financial products to improve lives.

Read the entire interview here.
black_portfolio-08.svg
Quantitative & Asset management

Companies in this category are largely employing AI to either improve the predictions or to create new funds based on algorithmic trading – so called quants. The company Trumid is an all-to-all electronic bond-trading platform applying AI to increase liquidity and transparency in the market for corporate bonds. Wealthfront is an example of a company offering a (b2c) product focusing largely on the creation of a new customer experience within AI-enabled asset management.

Also in the Nordics, traditional financial institutions such as Nordea are developing a new big data and AI-enabled credit security business. Scroll down to read all about it.
black_document-20.svg
Business Finance & Expense Reporting

Another significant example of AI-based horizontal services is found in business finance and expense reporting, where machine learning can be utilized to identify patterns and spending context to ultimately discover fraud.

The company AppZen provides an AI-based SaaS solution that enables real-time analysis of expense reporting for an in-time analysis of potential fraud. Their solution is pre-integrated with some of the existing reporting systems, and by utilizing AppZen, users become empowered to prioritize resources on pre-selected cases noted as potential fraud.  

During the Applied AI Project, the CEO of AppZen, Anant Kale shared his knowledge with Innovation Centre Silicon Valley with the purpose of investigating how AI can be applied to make an impact in the financial sector in Denmark.

 

black_big_data-46.svg
General purpose & Predictive Analytics

Applying AI to real business problems and achieving solutions resulting in tangible value is the ultimate goal for companies. However, for many companies, AI has until now earned a reputation as "something going on in the Data Science Team". The company Ayasdi is tackling this with a philosophy of intelligent applications, and how such end-to-end solutions can empower subject matter experts and business leaders with extraordinary capabilities. Through the Ayasdi AI platform it is possible to design, develop and deploy enterprise-size intelligent applications.

The Swedish founder of Ayasdi, Gunnar Carlsson, Stanford Professor Emeritus, is one of the most renowned mathematicians in the world. Innovation Centre Denmark Silicon Valley had the pleasure of interviewing Gunnar Carlsson in November 2017 about his perspective on the increased focus on AI from both companies and the public.

Read the full interview with Gunnar Carlsson here.
black_security_2-26.svg
Regulatory, Compliance & Fraud Detection
The use of AI to comply with regulations and detect fraud is a popular use case. Paytm - a world-leader within payment solutions, which have recently raised $1.4 BN (one of the largest funding rounds in the history of FinTech) - is an example of such. By applying their AI machine learning technology to payment transaction data the company is able to identify fraudulent transactions that would otherwise not have been detected.
black_security-27.svg
Insurance

Being able to offer quotes and insurance more intelligently are some of the prime examples of applying AI in insurance.

A company such as Lemonade is not only utilizing artificial intelligence to replace brokers and bureaucracy with Machine Learning and bots offering renters and homeowners insurance, but they also donate to non-profits.

In this way, AI is not only used for optimization of existing service offerings but is also enabling a new customer experience of remaking insurance as a social good.
black_trend_up-43.svg
Market Research & Sentiment Analysis

With the aid of sentiment analysis, it has become possible to enrich value analysis and evaluation with a new dimension.

An example of this is the company iSENTIUM, who provides sentiment data through Natural Language Processing, NLP, of social media feeds supporting 8.500 U.S. stocks, indices, and EFT's.

As a result, sentiment data can become an indicator of price movements. iSENTIUM allows real-time analysis as well as analysis of historical data thereby empowering users to create their own signals, trading rules, and strategies.
black_loan-16.svg
Debt Collection
Debt collection has long been a costly affair, but AI-enabled solutions for automated personalized communication can optimize the likelihood of successful collection. Companies such as CollectAI deploy an end-to-end solution, which automatically and intelligently uses the most successful communications channels, language tone, and time of contact for each bill, personalizing and automating the overall communication.

What AI solutions will transform banking and finance?

Working through the many different use cases, we have collected six sample companies with AI solutions that are being applied today in the banking and finance space transforming the sector: Vouch Financial, AppZen, Affirm, Ayasdi, Paytm, and Ant Financial.

Nordea logo.pngBeing among the 10 largest universal banks in Europe in terms of total market capitalization, Nordea is a force to be reckoned with.

Tracing its roots back to over 300 banks across the Nordics, some dating almost 200 years back, Nordea saw the light of day from the merger of Nordbanken, Unibank, Christiania Bank og Kreditkasse and Meritabank in 2001.

The name comes from a combination of the words Nordic and idea signifying how the organization seeks to share and develop good Nordic ideas to create high-quality solutions based on common Nordic values such as openness, equality and caring for the environment.

Highlights of The Road Towards The Future

2015

  • New CEO signals the beginning of a new era. An accelerator programme for FinTech start-ups to accelerate their ideas together with Nordea, is born.

2016

  • Nordea establishes new digital unit (Group Digital) as part of their ambition of becoming a truly digital bank.

2017

  • Formal partnerships with FinTech hubs in the Nordic capitals to incubate and hatch innovative solutions.
  • First consumer product from partnership. Nordea Liv and Norwegian FinTech start-up Spiff enter into a cooperation agreement to develop a social savings app for the user to set aside money for a trip, future dream home or pension with a few keystrokes.
  • IT innovation award. Nordea wins the 2017 Global Retail Banker IT Innovation of the Year award for their Simplification program (part of One Nordea) aiming to consolidate common processes, products and systems across the organization to create a truly digital bank.
Nordea at a glace: Key facts

Key facts.png

 

Nordea Asset Management: Developing a New Big Data & AI enabled Credit Securities Business

  • Project and team inspired from Exponential Finance to do a Nordea “Moonshot"
  • Decision to base it on the principles from Lean Startup
  • Purpose of significantly increasing assets under management at a dramatically lower cost
  • Focusing on three fundamental elements of new skills, data and computer power. 
MTP.png

Project initial focus: Opportunities in the process of selecting of credit securities for customers

Some challenges include:
  • All relevant data not organized and accessible in one system
  • Manual and discretionary environment using mainly traditional office type IT systems
  • Inability to analyze all relevant credit securities with the same depth
  • Automatically linking short term news flow from various news sources, combined with sentiment analysis to enable valuation of the credit securities.
Measure, learn, build.png
  • Move from fundamental to quantamental analysis
  • Reduce the reliance on gut feeling and ultimately develop AI enabled pricing prediction

Issues and themes

AI models within the asset management business not standardized in the industry
  • Data, cloud and market regulation
  • Alignment of organizational interests and structures
  • Technology inspiration and academic partners
  • Need to design, develop and implementation internal models
  • A critical step in the development of the solution is to set-up an independent and agile team with a diverse skillset hired from the outside except the domain expert who previously lead the existing credit securities business
  • Team works independently of existing business

Team.png

Cloudera

An example of a Big Data and AI solution for the enterprise is Cloudera. 

Other relevant companies besides the large global IT providers include Cloudera (above), Hortonworks, Ayasdiand  Palantir.

Selected Research Case in the AI Banking and Finance space