JP Morgan's recently released 280-page report Big Data and AI Strategies – Machine Learning and Alternative Data Approaches to Investing paints a picture of a future in which alpha is generated from data sources like social media, satellite imagery and machine-classified company filings and news releases. Alpha generation has always been about information advantage: having access either to uncommon insights gained through ingenuity, or common insights acted upon before everyone else. JP Morgan's Contract Intelligence System processes the paperwork for financial deals that previously took tens of thousands of human hours annually. Retiring old systems and moving to integration and data-centricity will require investment and some decent amount of vision, but it will result in future opportunities and cost savings: both from automation and from the ability of such systems to better take advantage of rapidly accelerating advancements in AI, which will require smart data collection, processing and management.
JP Morgan has a program that relies on artificial intelligence and that can save lawyers 360,000 hours every year (1). The program used by JP Morgan is called COIN (short for Contract Intelligence) and it does what its name suggests: read and interpret contracts, more specifically commercial-loan agreements. In order to achieve that, a technology called Machine Learning was used. COIN helps the bank save money and time by allowing the lawyers to work on tasks that are less redundant and more important.
Briefly, deepPiXEL provides AI solutions focused on text-based conversations for businesses in finance, retail and telecom industries. It is estimated by IBM that Decision Support will create $2 trillion worth of IT spending by 2025 (beyond the $1 trillion companies already spend on software, services, and hardware). With the help of AI, companies can improve the customer experience, augment employee performance, automate work processes, and develop intelligent agents to help with a lot of repetitive business processes (as JP Morgan has started doing). It's important to note that compelling evidence is already showing that investing back into office AI solutions provides the greatest return among consumer facing solutions.
Briefly, DeepPiXEL provides AI solutions focused on text-based conversations for businesses in finance, retail and telecom industries. Data of all sorts will be used to understand, reason, talk, make decisions and learn. With the help of AI, companies can improve the customer experience, augment employee performance, automate work processes and develop intelligent agents to help with a lot of repetitive business processes (as JP Morgan has started doing). It's also important to note that compelling evidence is already showing that investing back office AI solutions provides greatest return among consumer facing solutions.
In news more suited to the likes of Amazon, JP Morgan is set to look at customers' spending histories in order to sell them future products or services. Although AI is predicted to change whole sectors in finance (from asset management to compliance), I've always thought that the first applications of AI in banks would be much less spectacular – but much more useful – and that would be the tried and tested recommendation engine a la Amazon. CBR and FT reported yesterday that JP Morgan is launching a smart system to assist their sales people and suggest trades to clients – in other words, a recommendation engine for the trading floor. My guess is that we will start to see more and more banks applying similar technologies – not only because it's the normal evolution of the sales/client relationship, but also because they wouldn't want to see JP Morgan have a competitive edge… I've been saying for years that the AI needs to come to banking through customer relationship and I'm glad JP Morgan is showing the way.