Read More

Being human—AI and reasoning in transaction management

07 SEPTEMBER 2016
By Bill North

Why the ability to put information into context is at the heart of ‘next generation’ payment platforms. Artificial intelligence (AI) has been around for decades in various forms, but what is it really? Well, it’s the ability to put information into context.

Years ago, an internet search for “shoes” would yield millions of hits, including everything from trainers to Jimmy Choos to horse shoes—there was little intelligence involved. Refining the search—using symbols in the search bar—got it down to fewer results. Now? We get an accurate listing by relevance, even based on our own search history and preferences. What changed? The answer lies in context. When there is context  we can do much more with data. We can understand what information means and make more relevant and intelligent decisions. For example, a search engine may know that you like running, so it starts by offering you trainers. If it knows your location, it may suggest local shops. Context makes things relevant. Humans use context every day—it’s fundamental to our intelligence, our understanding. We call it common sense or general knowledge.

Payments for the ‘I-want-it-now’ generation
Most financial institutions base their business on ‘first generation’ payments systems, which have improved straight-through processing, especially compared with manual processes. However, that approach is rigid and unintelligent. Specifications—for example formats—must be followed precisely, and any deviation is rejected and handled manually, at great cost. In addition, there is a focus on the individual payment, with little ability to see or understand the entirety of activity. And there is very little scope for innovation—almost all attention is directed to basic capabilities. People talk and understand in natural languages like Spanish, English and Japanese, whereas computers talk using codes in bits and bytes. We can use technology to bridge this gap, and make computers work and understand more like we do—by giving them some common sense.

‘First generation’ payment systems may be well established, but they cannot offer the adaptability or ease of use people have come to expect, especially the ‘I-want-it-now’ generation. Nor can they easily change to meet the challenges of globalisation and regulatory developments—not to mention customers’ expectations for innovative services. Finally, commercial pressures have driven down the acceptable cost per transaction, which, together with increased competition—also from non-traditional players—puts a strain on budgets that especially affects areas like payments that are viewed as cost centres. The diagram summarises the evolution and shows the possibilities.

Pelican: payments... but different
Next Generation’ platforms like Pelican approach the challenge differently—for example, by using AI-based ‘human’ general knowledge to read and understand payment messages the way a person would, and use context. This approach allows Pelican to:

- Provide visibility of all activity across the entirety of a business
- Repair or re-format or enrich a payment based on context and the client’s intentions
- Make informed decisions on the quickest, most optimal or least cost route for the payment
- Provide true protection for compliance officers, and provide details on decision

AI capabilities could of course be used to handle payments ‘old style’—a bit like using a laptop to print an email, then posting it. The point is, a next-gen system contains the architecture to allow companies and financial institutions to improve their payment capabilities incrementally by simply plugging in required components when required—for example, by using payments data to begin managing liquidity much more effectively, with an on-going intra-day picture, via configurable dashboards and reports.

A next-gen system will also provide the flexibility to use its capabilities without the need to install, run or maintain software. The advantages of cloud computing are well-known—and SaaS is finding increasing appeal with companies and financial institutions. Similarly, a modern platform allows both corporate treasury and banking professionals to initiate, monitor and control transactions while on the move. A “mobile” interface has become a requirement.

Pelican for transaction banking
Pelican’s strengths start with simplifying and automating connectivity—communication protocols and the structure of the content don’t matter. Organisations greatly benefit from using a standardised gateway that provides efficient and secure connections, and takes in payment messages from any source: banks, clearing systems, a SWIFT gateway, an FTP or other connection, an electronic banking server, or any other systems or communications channels.

Pelican is available on one extensible platform, with no need to tie together or connect different pieces of software. Capabilities can be added as needed, and the platform easily integrates with existing systems, allowing you to build a comprehensive end-to-end solution. Pelican is scalable and uses memory-based processing to guarantee the high throughput required for even global organisations with very large volumes.

Context and intelligence in sanctions screening
The Pelican platform’s intelligent approach comes into its own in the world of sanctions screening, where context and understanding is more important than rules: rules set limits or constraints, whilst context is something humans apply to understand a scenario or piece of information, and make intelligent decisions. Use of contextual intelligence drastically reduces ‘false positive’ rates and significantly reduces the costs associated with reviewing false alerts.

The focus on regulatory compliance in wholesale payments has grown significantly. Governments are not only looking more closely at sanctions compliance, but are looking in much more detail at how companies and banks set up and execute their compliance programs.

Heightened scrutiny has spotlighted the need to filter accurately, but business pressure means there can be no loss in operational efficiency. A truly effective screening tool understands context, to minimise the “noise” of false readings—and can even play a role in building competitive advantage. The ability to provide detailed and complete information on all positive and negative decisions demonstrates competence and builds confidence with regulators, clients and shareholders.

Pelican: the right approach
The ability to put information into context and use it intelligently is at the heart of ‘next generation’ payment platforms. Pelican is scalable, flexible and easy to integrate, and uses contextual understanding to provide a range of value-add services that traditional systems don’t.

It’s time to look at how you process and screen payment transactions, and how the right approach can reduce costs, create efficiency, and build understanding of your business. We call it common sense.

Back to Articles