System integration brings data and workflows together, allowing organizations to work more efficiently, improve collaboration and increase productivity.
In this episode of Cherry Bekaert’s Digital Journeys podcast and the second part of our system integration series, Jim Holman, Director and Strategy & Operations Leader, welcomes Jon Schrader, Director and Business Applications Leader, as they summarize the various layers of system integration.
Listen to learn more about:
- Different types of system integration architectures to best suit your business needs.
- Explanation of integration platform as a service (iPaaS) consists of a suite of cloud-based services.
- Point-to-point integration involves custom coding, application program interfaces (APIs) or both.
- Robotic process automation (RPA) mimics how people interact with software and workflows.
- Common data integration tools and how they correspond to other software to empower businesses to reach their full potential.
Stay tuned for part 3 of this series, where we will further discuss the importance of system integration.
If you have any questions specific to your situation, Cherry Bekaert’s Digital Advisory team is available to discuss your situation with you. Contact us today!
Related Insights:
View All Digital Journeys Podcasts
HOST: Hello, everyone, and welcome to Digital Journeys. Today, in part two, we are joined by Jim Holman and Jon Schrader, who will discuss the importance of system integration. With that, I turn it over to Jim.
JIM HOLMAN: Thanks for joining me today, Jon, in part two of this series.
JON SCHRADER: Great to be back with you, Jim.
JIM HOLMAN: You had mentioned integration tools before, and I know that there are a lot of acronyms and different types of tools. Can you take us through the common types of integration tools available and how they connect to current software systems?
JON SCHRADER: System integration can occur at multiple layers. Software systems typically use databases to store data, and traditionally integrations were built at this database layer by writing queries or stored procedures to extract or insert data directly into the system database.
JON SCHRADER: Direct database integration is dangerous and tedious because it can impact the operational stability of the system and create security concerns by allowing deep access to data. Most modern software provides application programming interfaces (APIs) to get data in and out, and API-based integration is the preferred method because the software exposes how it wants you to interact and maintains its business logic.
JON SCHRADER: APIs are typically exposed through secure web services, and that is our preferred method of system integration. If API-level or database-level integration is not available, integration can occur at the user interface layer using automation tools that mimic a user or send keyboard commands.
JON SCHRADER: UI-level integration is done using robotic process automation (RPA). RPA has been a recent buzzword, but this type of integration is unstable and should be a last resort because user interfaces change frequently and can break UI-layer integrations.
JON SCHRADER: There are different categories of integration solutions. Integration platforms as a service (iPaaS) are subscription-based, cloud-based, API-focused platforms that offer benefits such as out-of-the-box connectors, data transformation, duplicate elimination, and support for citizen development in data mapping.
JON SCHRADER: iPaaS platforms can perform extract, transform, and load (ETL) functions and allow non-developers to map data using low-code or no-code tools. Some platforms, such as WorkAuto or MuleSoft, also provide replay capabilities for integrations and transaction failures.
JON SCHRADER: The next category is point-to-point or custom-developed integrations using API or database approaches. The problem with point-to-point integration is that you end up with different solutions for each integration and lack a common language across systems.
JON SCHRADER: Finally, as mentioned, robotic process automation is a UI-layer integration and should be used only when API or database integration is not possible.
JIM HOLMAN: I have had direct database integrations in my career that went very badly. They become vulnerable to system upgrades and unexpected changes that cause integrations to misbehave, and it can be difficult to realize the extent of the damage they can cause.
JON SCHRADER: I had an experience where database changes rendered the entire ERP platform unusable and crashed the sales process. For reasons like that, database-layer integration is not recommended.
JIM HOLMAN: You mentioned citizen development. Some of these iPaaS platforms are low-code or no-code and enable business analysts, rather than developers, to drag and drop fields between connectors. Many iPaaS platforms include out-of-the-box connectors.
JIM HOLMAN: For example, if an iPaaS platform supports Shopify and Dynamics 365 Business Central, mapping a customer from one system to the other may not require coding. It might only require light scripting if you want to modify the customer identifier during the transfer.
JIM HOLMAN: These platforms create opportunities to reduce dependence on IT specialists. Thank you for your time today, Jon. In part three, we will look at other important areas of system integration. Thank you, Jim and Jon, for the discussion today.
HOST: We appreciate your insights on system integration. Feel free to like and share this podcast. Stay tuned for part three of this series.
