Page 35 - gs261002
P. 35
Insights and Expertise
For example, identifying a reconciliation exception may Rather than:
require more than comparing two amounts. It may re- AI — Hope it fixes everything
quire understanding the original transaction, associated
fees, settlement activity, currency, adjustments and other The back office is an AI opportunity
attributes. Likewise, investigating a dispute may require
connecting the dispute to the original transaction and the
activity surrounding it. I can be extremely helpful in ana- None of this means payment organizations should wait to
lyzing this information, but first, the information needs explore AI. Quite the opposite. The back office may be one
to be connected. A back office that maintains a connected of the most promising areas for practical AI applications
view of the transaction lifecycle provides a much stronger because it contains enormous amounts of structured
foundation for intelligent automation than one where in- transaction data and numerous processes that involve
formation is isolated in functional silos. research, analysis, matching and decision-making.
Don't put AI on top of bad processes Potential applications include:
There is another temptation when organizations adopt • Reconciliation and exception management
new technology: automate the process they already have. • Transaction research
If employees manually review thousands of exceptions,
the instinct may be to use AI to review those exceptions • Fee assessment
faster. But perhaps the more important question is: Why • Settlement processing
are so many exceptions being generated in the first place? • Disputes management
If reconciliation requires excessive manual intervention, These are all areas where intelligent automation could
organizations should examine the underlying matching potentially create significant value. But the organizations
rules, data quality and process design. If employees spend that benefit most may not be those that simply adopt the
hours researching transactions across multiple systems, newest AI technology. They may be the organizations that
AI might help them search faster, but a centralized trans- first create the operational foundation that allows AI to
action repository and better lifecycle visibility could elim- work effectively.
inate much of that research in the first place.
The foundation matters more than the hype
Similarly, if fee management requires extensive manual
analysis, AI might identify patterns, but flexible and con- Every new technology cycle creates pressure to adopt the
figurable fee rules can prevent many problems before they latest innovation. Today, that innovation is AI, and there
occur. AI should not become a substitute for process im- is good reason for the excitement. But payment process-
provement. It should become an extension of it. ing is too critical to approach technology as a collection of
disconnected solutions. A payment organization can have
The modernization sequence matters sophisticated AI capabilities and still struggle with basic
operational problems if its transaction data is fragmented,
For payment organizations considering AI, modernization its processes are overly manual and its systems cannot
doesn't have to mean replacing everything at once. But provide a connected view of the payment lifecycle.
there is value in thinking about the sequence:
1. Connect the data. Create a consistent view of the in- AI doesn't eliminate the need for a strong payments back
formation needed to manage the payment lifecycle. office. It makes having one even more important. The fu-
2. Improve the processes. Identify unnecessary man- ture isn't about choosing between a modern payment in-
ual steps, disconnected workflows and sources of frastructure and AI. It is about building payment opera-
recurring inefficiencies. tions where the two work together. The organizations that
get the most value from AI will likely be those that recog-
3. Automate what makes sense. Use technology to nize a simple principle: Before you make your payments
eliminate repetitive work and reduce opportunities back office intelligent, make sure you make it connected,
for error. visible and operationally sound.
4. Apply intelligence. Once reliable data and effective Michael Meeks, chief technology officer for BHMI, is responsible for the
processes are in place, AI can help organizations company's strategic technical development, software development
identify patterns, make recommendations, priori- standards, and software management procedures. BHMI is a leading
tize work and support decision-making. provider of software solutions focused on the back-office processing of
electronic payments. The company is best known as the creator of the
This creates a stronger progression:
Concourse Financial Software Suite, an integrated collection of back-
Connected data — Better processes — Automation office products that allow companies to adapt to the rapidly changing
Intelligence world of payments. Michael can be reached at mike.meeks@bhmi.com
or at 402-333-3300.
35

