The developer and IT skills gap is not a new problem. But because business demand for new technology is increasingly multifaceted and complex, the challenge of finding enough skilled employees is only going to get tougher. Organisations needing specific knowledge have been limited to one of two choices for a long time: hire in the people (directly or by contract), or don’t do the work — until recently.
AI is the new kid on the block, presenting a third option. Its appeal to many people, governments and businesses is that it can fill the [skills] gap for some specific tasks. And while one AI system can not, and will not, replace a skilled developer, finding specific, specialist tools to automate some functions will allow developer and IT teams to operate more broadly and effectively.
The future impact of artificial intelligence — This article will explore how artificial intelligence is set to impact organisations in the future, gauging the insights of experts in the space.
The challenge now though is to really define what benefits AI can bring and how it can be effectively integrated into existing operations. It’s not replacing people, rather working alongside them… so what does that relationship look like and how does it work?
When it comes to the task of developing integrations, there are several layers of complexity. Not all integrations are the same; some are relatively straight-forward, while others can take large teams and many months. Nonetheless, even the ‘simple’ ones take time. There are several stages of development required before introducing a new integration, starting with task definition, building glue code, testing, and maintenance. If AI is to help in this area, it needs to master multiple tasks.
Every business needs more integration
Businesses have been on a transformation journey to become more digital for years. There are more connections than ever that help us with our day-to-day tasks, and there is an essential need to move data between systems to unlock value. The vital aim for companies today is finding how they can get more value from their data and integration is the key.
The aim for the future is give all employees the tools to build their own basic integrations with minimum technical support. Imagine if Sarah from finance could create an integration between order processing input and a payments platform on her own? And if Nick from HR could generate an integration between employee feedback via email and Workday?
One-time integrations are extremely valuable, and can impact the productivity of individual employees and/or small teams. However, encouraging teams to do this now would create too many requests for IT teams to handle. Simply put, there aren’t enough skilled developers in the world to meet demand. The answer is in AI — more specifically an autonomous AI integration specialist that anyone can access through a natural language chat interface. While that might sound unlikely, we are already on our way towards that goal.
Optimising a team of AIs
Even though there’s currently not one AI that can create all the layers of development required to code a new integration from scratch, we don’t actually need it to reach the end goals. The process can be completed by getting together the right ‘team’ of AIs under a straightforward chat interface, drawing influence from programmes like Auto-GPT.
The future of AI is not a uniform environment dominated by a single, superior tool. Rather, we’ll likely see AI specialists that are extremely effective and precise when performing specific tasks. Businesses will derive value from the extremely wide concept of artificial intelligence in this way. In order to take advantage of the capabilities of multiple services, multi-AI will become just as crucial as multi-cloud support, which is a major feature of today’s integration stacks.
The question of how this diverse blend is produced and subsequently utilised then arises. The solution is using one AI-enabled integration layer to manage all the others; crucially a layer that’s driven by a conversational interface, in the same way that ChatGPT has proven so popular.
This one AI can be trained to coordinate all of the other tools needed for the process of building integrations. The process needs, amongst others, one AI to create a task list; another to go out and find the existing API code; another to write the new code; then another to test and manage. All these tasks are completed by this team of ‘AI engineers’, under the supervision of the primary tool — such as ChatGPT — which serves as both the interface and the project manager. The AI can be taught to query the responses to important steps in the process, for example. The user can simply accept the result, or they can ask questions and request changes or updates as needed.
84 questions to ask before training an AI model — AI experts and data scientists have published a checklist of questions developers should ask before embarking on training an AI model.
Having this capability is huge and we’ll likely see it most used in two main business cases. Firstly, the ability to democratise integration building for the business, making it quick and easy to integrate information. The speed this works at is also significant: staff will barely have time to make a cup of tea while AI is working on the build. Thus, this wholly improves the productivity of employees within a business, and hopefully reduces stress levels too.
The next benefit is the productivity of a company’s limited developer resources. Developers can concentrate their time more effectively and some tasks will cease to exist at all. For more complicated integration, success will require that expert human eye to collaborate with the AI platform and this process will be substantially more difficult and involved. However, “simple” integrations can be completed quickly. By “simple”, we mean those where the time required is frequently spent locating the APIs from platform developers and obtaining the end users’ briefs, but the actual coding is not too complicated or out of the ordinary.
Naturally, more work will be required to maintain the AI itself; in spite of their apparent lack of code, no code systems actually call for more complex “minders”. Another way to free developers from several necessary but boring chores is to use AI to eliminate some forms of technical debt.
The focus has now shifted away from “if” AI will be a tool for commercial value, and towards “how” and “when.” The trick lies in using everything that is already in place. It’s not about attempting to force a square peg into a round hole with an AI unsuited for the job at hand, or attempting to develop every integration from scratch. The companies of the future will have AI specialists dedicated to increasing productivity, improving employee happiness, and understanding the value of data.
AI could add $500bn to value of SMEs globally — SMEs either creating AI for enterprise organisations or embedding AI in their own businesses will add billions of dollars in value, says boutique tech bank DAI Magister.