The Secret Path To Practical AI

8 minute read | 21 Nov 2023

By David Irecki

The hyper-buzz around AI’s potential to transform business has left many organizations facing some vexing questions. The biggest question is a simple one: What should be our first step?

The options are many. You could experiment with ChatGPT, the fastest-growing consumer app in history. You could customize your own large language models. You could hire a couple of AI rock stars to help guide your journey.

There’s a better approach — one that gives you a flexible and resilient foundation to take advantage of AI’s potential for years to come.

At Boomi, we call that approach “practical AI.” We can define practical AI as the use of AI technology with the laser-focused goal of driving business value in areas such as:

  • Delivering engaging experiences for customers, partners, and employees
  • Streamlining and optimizing core business processes through automation
  • Innovating products, services, and even entire business models

You may be thinking: Aren’t those objectives pretty much the same as what we’ve been after with digital transformation?

Practical Steps You Can Take Today

The answer is yes — but AI has changed the equation so dramatically that businesses around the world are scrambling for ways to capitalize. Standing still is not an option, unless your business is willing to forfeit competitive advantage to rivals.

Asia-Pacific businesses are especially bullish on AI’s potential. Nearly 90% of CEOs in APAC say their companies have made or will make significant investments in AI technology within the next 12 months, a recent Ernst & Young survey found.

That aligns with separate research studies that predict Asia-Pacific will be the fastest-growing market for AI among all global regions, as I highlighted in a previous blog post, “Asia-Pacific Poised for World-Leading Growth in AI Adoption.”

The blog post also notes that about 75% of large Asia-Pacific organizations will use AI-infused processes to boost asset efficiency, streamline supply chains, and improve product quality by 2026, according to a report by IDC, a leading global consultancy.

So what’s the secret path to driving business value with practical AI? Let’s review how the enterprise landscape has evolved, how AI stands to disrupt the status quo, and practical steps that you can take today.

Bridging the Disconnect Between Insight and Action

Enterprise infrastructure has traditionally been anchored by two types of systems — systems of engagement and systems of record.

We can think of systems of engagement as customer-facing web, mobile, and chat applications operating in real time. Systems of record are those more static ERP, CRM, HR, and other back-end applications that store and process mission-critical data.

With digital transformation, many companies are integrating data and applications, and automating workflows, across systems of engagement and record using technology like integration platform as a service (iPaaS).

Enterprises have also built out systems of analysis, using data warehouse and business intelligence tools. But many have found it difficult to connect systems of analysis with real-time systems of engagement in a timely manner.

In other words, systems of analysis might generate invaluable insights — but not quickly enough for those insights to help drive intelligent, real-time interactions with customers, partners, and employees.

That’s where AI comes in.

AI is turning systems of analysis into systems of intelligence that are better, faster, and smarter than contemporary analytic tools. As these systems start to operate in real time, the disconnect between insight and action can be bridged with integration and automation. And this is just one example of gains that enterprises can realize with practical AI and a flexible data foundation.

Practical AI Enablers: Context and Action Pipelines

Let’s dive in to what an enterprise AI landscape will look like from a data engineering frame of reference. As organizations have evolved their data capabilities in recent years, many have built sophisticated data pipelines to supply data to data warehouses and lakes.

Similar pipelines can be used in an AI landscape, where companies are using many types of AI models and tools. To generate relevant business insights, these tools require enterprise data to be put in context.

The context pipeline fine-tunes data generated by pre-trained AI models to optimize accuracy and help ensure relevance of insights. Once those insights are enriched, they need to be brought to the forefront of the business as quickly as possible.

That’s where the action pipeline comes in. The action pipeline gets the insights into customer-facing systems of engagement, informs customer service personnel through their systems, and helps automate core business processes that drive the company’s bottom line.

What used to be a long manual process — data analysts share findings with various teams, projects kick off to act on the insights days or weeks later — can now be done in near real time through AI, and delivered in real time for execution across systems of engagement through the action pipeline.

Are You Already On the Secret Path?

How can organizations construct these pipelines that are foundational to practical AI? What is the secret path?

The good news is that you can start creating these pipelines before you start explicitly adopting AI. In fact, you have probably already started!

For the context pipeline, you need capabilities including data integration, automation, and data governance. If you have an iPaaS like the Boomi platform in place, you have a huge head start on building your context pipeline. You’re able to enrich data with context, ensure its completeness and accuracy, and automate data flows to target destinations.

Capabilities required for the action pipeline are similar. Just as you need integration to source and synthesize contextual data, you need integration to connect insights with the frontline systems of engagement that execute core processes and orchestrate user interactions.

From a governance perspective, the action pipeline benefits from API management to help connect with the external ecosystem of model providers and ensure that data privacy is maintained. You also need automation to have the AI-generated insights drive business value in real time.

Future-Proofing for the AI Revolution

It’s also important to think of how your organization can future-proof itself for the hyper-speed pace of change in the AI revolution. Knowing that the AI of today will be obsoleted and replaced in short order, it’s important to adopt a methodology for introducing change that takes advantage of your flexible new pipelines.

Three principles for delivery can help guide your journey:

  • Solve current pain points utilizing intelligent integration and automation; each solution to a current problem advances development of both context and action pipelines.
  • Break those problems down into smaller increments to get to value faster, limit impact of any issues, and drive more flexibility.
  • Relentlessly enforce a composable approach to solution architectures to ensure you can harvest meaningful digital building blocks at every opportunity.

Adopting these principles will create a balance in hitting near-term practical goals and paving the way practical AI that can adapt and scale over the long term.

Building Organizational Muscle for Practical AI

The secret path to practical AI is one that your organization may already be on, but it requires diligence around the approach and establishing the foundational capabilities.

By solving these problems in a composable, capability-led way, you will build up organizational muscle to power your company’s practical adoption of AI. You will create an array of digital assets, an army of enabled users from a broad spectrum of skills, and a greater level of visibility and control over your IT estate that becomes vital when looking for opportunities to innovate through AI.

Practical AI also benefits from a robust, dynamic data foundation. The intelligent integration and automation capabilities of the Boomi platform supply that vital data infrastructure, equipping you to accelerate value from your AI initiatives.

Are you AI-ready? Learn about Boomi AI, a first-of-its-kind tool that uses generative AI to intelligently integrate and automate across applications, data, processes, people, and devices, and take our free AI-readiness quiz to see where your organization falls on the path to AI adoption.