How Next-Gen iPaaS Architectures Are Delivering Intelligent Automation

Even with advances in low-code integration and app development, digital transformation projects can still lag.  Dell Boomi’s Ben Thrift looks at today’s architecture trends can speed things up – thanks to a healthy dose of automation and intelligence.

Tags: automation, Boomi, cloud, integration, intelligence, iPaaS, low-code, multi-tenant,

Ben Thrift, Dell Boomi
Ben Thrift
head of engineering 
Dell Boomi

"In 2019, the opportunities for intelligent automation are rich and more varied than ever. The key to success is to automate the creation of automation itself."

Application Architecture Summit
Modern Application Development for Digital Business Success
Online Conference

Despite IT advances such as mobile apps and cloud computing, organizations still don’t work as quickly and efficiently as possible. The culprit? A lack of automation. Too much daily work remains manual and piecemeal.


Workflow applications promise to fix this problem by automating processes, delivering data quickly and easily to people when and where they need it, and removing operational “friction.”


But so far, businesses have trouble developing and deploying all the workflow applications they need. Automation is typically limited to a few high-profile, mission-critical processes. Meanwhile, most of their internal processes remain unautomated, dependent on manual steps, paper forms, and calendar reminders to send a certain email to a certain user at a certain time.


What’s behind this automation backlog? A big part of the problem lies in how companies develop applications.


Use Intelligent Automation to Tackle App Backlogs

Most workflow applications are built through back-office development projects that last for months. The projects begin when business users meet with developers and describe what they need. Developers take these requirements, disappear for weeks to write and debug code, and return with a Minimum Viable Product (MVP) or perhaps a more advanced version of the software that the business users have requested.


Even if that first version of workflow automation software is beyond the bare minimum, additional iterations of development and testing are usually required.


Add it up, and the entire process from the initial interview to the delivery of a useful, reliable software application is slow, resource-intensive, and error-prone.


And while the result at the end of this intensive, multi-month project may be that a single workflow app has been automated – dozens more remain in the queue.


Below we’ll discuss a new approach for tackling a mounting automation backlog, where a new combination of tools empowers both seasoned app developers and non-technical business users.


Intelligent Automation with Low-Code, Next-Gen iPaaS and Machine Learning

Let’s look at a new approach for intelligent automation using low-code, intelligent iPaaS (integration Platform as a Service), machine learning, and terabytes of anonymized meta-data.


Many organizations have already enlisted low-code approaches to accelerate their workflow automation by enabling business users to do much of the development themselves.


In a low-code development environment, even non-technical users can produce working software using a drag-and-drop canvas of shapes.  They build connections, decision points, and workflows on the design pane of a graphical interface and create business value in real time.


Low-code development puts subject matter experts in charge of building their own solutions and driving their own priorities. It’s a big step forward to realize an organization’s automation backlog and make business processes faster, more reliable, and more efficient. Even if complex projects require some assistance from developers, that assistance consumes only a small fraction of the time traditional software projects require.


In fact, statistics show low-code development dramatically accelerates software production -sometimes by a factor of 10X.

But can the development of workflow applications and automation be accelerated further, while keeping both developers and business users on track?


Enter the “intelligent iPaaS,” a next-gen, multi-tenant iPaaS that can leverage huge metadata volumes from hundreds of enterprises applications and endpoints.


Business users can design and build workflow automation applications that connect key business applications, such as Salesforce and SAP, on a multi-tenant iPaaS platform running in the cloud. Because the platform is cloud-native, users can build, manage, and deploy their applications from anywhere.


Because the intelligent iPaaS platform is multi-tenant, it can constantly monitor and collect metadata (anonymized, of course) on millions of tasks performed by thousands of users. From that data, the intelligent iPaaS ‘learns’ the best ways to accomplish these tasks and automatically suggests solutions for time-consuming actions, such as mapping data sets from one application to another.


Note that this type of guidance is only available from multi-tenant cloud platforms. Even the most advanced single-tenant system will never draw upon the experience of thousands of other companies and tens of thousands of other users. Instead, its suggestions are limited to those inferred from a company’s own internal development projects.


Guidance makes things faster. Learnings-based automation (based on real data not merely guesses) further accelerates application development and reduces guesswork and errors.


Expand Automation of Workflows, App Development

Today, several application vendors promise streamlined application development. But most of these solutions require a company’s data to be imported into that vendor’s applications and data stores. In other words, automation only becomes possible by adopting a single vendor’s proprietary approach to manage data.


For most businesses, that’s simply not practical.


Fortunately, today’s intelligent iPaaS can support an open ‘best-of-breed’ architecture, where companies can build integrations, apps and workflows with data from thousands of enterprise sources without the need to import any data.


In 2019 the opportunities for intelligent automation are rich and more varied than ever.  The key to success is to look for a development platform that helps automate the creation of automation itself.


Look for a multi-tenant iPaaS platform that provides an easy-to-use, low-code development environment and draws on the experience of thousands of other companies building integrations, configurating data transformations, and automating workflows. With that power in hand, you’ll be well on your way.


Ben Thrift is Head of Engineering at Dell Boomi, where he is responsible for the overall development and hosting of the Dell Boomi product line.  He has been in enterprise software for 20 years across a wide range of roles -- developer, architect, director of software and operations, and executive leadership.  Ben has also co-authored several US patents supporting large scale software architectures and the management of distributed systems.