The use of AI is fast becoming the norm for business processes. As the hype around AI continues, costs still rise and staff and skills shortages still increase. Simply implementing AI for the sake of it is not a good business reason, but using AI to transform some key development processes to drive real business value is.
Here, Edward Funnekotter, Chief Architect and Chief AI Officer at Solace, gives his top five business transformations that would not be possible without AI.
Growth in AI adoption is no secret, but how tech budgets will shift in 2024 to accommodate it is front of mind for many organizations. Predictions from IDC show that IT spending towards AI will be quick and dramatic and will influence almost every industry and application. By next year, Global 2000 organizations will distribute over 40% of their core IT spend to AI-related initiatives, spurring significant increases in the rate of product and process innovations.
Extracting Business Value to Make Investments Worthwhile
In 2024, we will see more organizations seek real business value from AI – rather than buying into the gimmicky consumer-facing applications of AI already on the market such as large language models or GenAI for text or images. This year, enterprises need to hone in on their own AI priorities to establish the desired outcome for AI investment – particularly as additions can cause subscription prices to skyrocket.
The 2024 balancing act is upon us, and organizations need to weigh cost to benefit ratios that ensure business value is felt across the entire company.
Prediction 1: The Secret Weapon – Transforming the Analog Nature of Events into Analogue Digital Events
In 2024, organizations need to shift from purely data “generation” to data and decision “velocity”. This needs an event mesh, a network of interconnected event brokers that enables the distribution of events information among applications. Using an event mesh makes it possible to layer AI inferences on top of each other, adding intelligence incrementally.
To explain, let’s consider a practical application that merges AI with an event mesh to react to the analog nature of the world to process and respond to diverse inputs such as audio, video and human text.
Imagine a building incident manager that’s overseeing a company facility. There are guards present in the building, as well as security cameras to properly respond to events that happen on the premises. But there are also bystanders and employees that have an ability to text or slack critical information to the building operations manager.
AI agents, each performing a specific role, can subscribe to the events from these inputs in order to transform them into more consumable events. Inside this mesh, those inputs will be taken and passed to the appropriate model based on a per-event hierarchical description of the event. So, let’s say the security guard radioed in that there’s a problem in the reception area. That’s useless to the incident manager application. However, if you push it through the speech to text model and augment that event with the actual text that the security guard used, it will then get routed to the incident manager who now has a text description of what the guard was saying.
It could then use the AI large language model (LLM) to decide what the appropriate action should be for the issue at hand. Some of those actions might need to be verbally spoken, in which case it can be directed to the text to speech model, which could then be played on the security guard’s walkie-talkie. Of course, this information gets augmented and passed through this incident manager, which is all AI driven. It could then publish into the event mesh to automatically trigger next steps, such as turn on alarms, turn off alarms or alert emergency services to take care of whatever incident might have happened.
The importance here is that an event mesh, combined with AI, translates a series of analogue events into a streamlined flow of information, allowing an extremely quick time to gain intelligence. As the example shows, the business value can be huge.
Prediction 2: Productizing AI
2024 will see AI and data getting “productized” into Intelligent Applications to deliver real business value and intelligent insights. Gartner defines intelligent applications in its strategic IT trends for 2024 as consumer or business applications that are augmented with AI and various connected data from transactions and external sources.
But in an AI-Everywhere world, data is the crucial asset to feed AI models and applications. There’s a data grab on right now.
Two challenges in the way of the great data grab – silos mean data silos!
Technology suppliers and service providers recognize this and will accelerate investments in additional data assets to improve their competitive position. This is well needed, as IDC finds only 12% of enterprises connect customer data between departments and 42% of enterprises have underutilized data.
And it needs Unified Control.
For AI and data to work together, it needs a connected enterprise and Unified Control. IT teams in the next several years will need to start navigating the maturation of control platforms as they evolve from addressing a few basic systems to becoming one platform that orchestrates operations across infrastructure, data, AI services and business applications and processes.
This means putting data in motion – to direct the right data to the right places and get it anywhere in the world.
Prediction 3: AI Tinted Glasses are in Fashion to Shorten Software Development Lifecycles
2024 will also present a growing opportunity for AI usage in application development. To what degree is for debate. Forrester, for instance, predicts generative AI bots, or TuringBots as they call them, will play a substantial role this year in shortening software development lifecycles by 15 to 20%.
In my view, it will be some time before development teams are fully embedding AI bots in their software development lifecycle. But the spirit is there – there is no question that AI technologies, such as generative AI and machine learning (ML), will aid software engineers in creating, testing, and delivering applications, providing an assistive role to help accelerate development tasks.
If anything, I believe 2024 will see development organizations spend more time looking at their software development lifecycles through AI tinted glasses, seeking to better gauge where current processes are flexible enough to embed AI in a way that provides real value. For example: AI-augmented development tools integrated with an engineer’s development environment to produce code, translate legacy code to modern languages, enable design-to-code transformation and enhance application testing capabilities.
Prediction 4: Introducing the Platform Engineer!
But such applications will be only hypothetical without the ability to design and develop them in the first place. This is why it is exciting that Gartner sees topics such as Platform Engineering coming of age in 2024. Platform Engineering is an emerging trend intended to modernize enterprise software delivery, particularly for digital transformation. It’s an approach that can accelerate the delivery of applications and the pace at which they produce business value.
It improves developer experience and productivity by providing self-service capabilities with automated infrastructure operations. It involves discipline of building and operating self-service internal platforms — each platform is a layer, created and maintained by a dedicated product team, designed to support the needs of its users by interfacing with tools and processes. The goal of platform engineering is to optimize the developer experience and accelerate product teams’ delivery of customer value.
Gartner predicts that 80% of software engineering organizations will establish platform teams by 2026 and that 75% of those will include developer self-service portals.
Prediction 5: Integrated Enterprises are Here – and Here to Stay!
But amid the rush to real-time and AI-driven operations, large, disparate organizations will still be limited in their ability to achieve optimal business value because of their reliance on a complex mix of legacy and/or siloed systems. Remember the IDC stat that only 12% of organizations currently connect customer data across departments! Constellation research agrees, stating, “Few enterprises have their data games down.”
This is why I believe the AI Data rush will drive greater industry-wide urgency for event-driven integration. This entails the combination of data transformation and connectivity attributes of an iPaaS with the real-time dynamic choreography of an event broker and event mesh. Only with this enterprise architecture pattern will systems new and old be able to work together to offer seamless, real-time digital experiences, linking events across departments, geographies, on-premises systems, IoT devices, in a cloud or even multi-cloud environment.
Time for a beer?
Consider an organization the size of international brewer HEINEKEN – which runs thousands of business-critical applications across 190 countries. In order to achieve its aim as “the world’s best-connected brewer”, the company introduced its EverGreen business strategy, underpinned by a shift to event-driven integration. In the past, HEINEKEN would see hundreds or thousands of point-to-point scenarios, but now they are being leveraged with one-to-many integration patterns, where an application only has to produce an event (such as an order of beer) once, and any other applications in the system (production, shipping, fulfillment, inventory, payments, cloud data lake) can just subscribe to what they want to receive, and get it when it’s published. Delivering seamless digital interactions across the entire value chain, HEINEKEN has now positioned itself to make smarter, more informed and real-time decisions – an organization truly getting on top of its data game.
AI and Business Value Create New Opportunities When Done Right
2024 is certainly the year of the balancing act for enterprises looking to navigate the land of AI. Of course, we will see AI integrated as part of new products and solutions at the front-end, but without data in motion, these apps won’t live up to the hype. To power the back-end, AI has the potential to drastically shorten software development processes, but there is still some headway to be made before we see this become a true reality.
AI-enabled or not, API and integration strategies are certainly on the priority list for an event-driven refresh, as it allows organizations to optimize operations at scale, and in real time.