Auto-tech series - SoftwareAG: Practical steps for developers integrating generative AI

This is a guest post for the Computer Weekly Developer Network written by Subhash Ramachandran in his role as senior vice president for product management at Software AG.

Reminding us that generative AI tools like ChatGPT and Bard are creating a lot of interest, Ramachandran says that there are still many unknowns about how they will be used, by whom, and for what. 

However, he notes, the technology offers as much promise as it does risk. 

In his team’s experience at SoftwareAG, generative AI can be most effective if deployed using the ‘80/20 rule’ of workload to resource. This is because so many tasks within business processes today are repeatable and do not require human resource: the perfect environment for automation.

The suggestion from Ramachandran is that is a technology evolution that can help bridge productivity gaps that are caused by the technical skills deficit. To help drive growth, businesses can hand over larger volumes of processing tasks to generative AI. In the right situations this can speed up and increase the output from these basic processes without draining valuable human capital.

Ramachandran highlights two real-world areas where generative AI is making a difference for software engineers and integrators and writes in full as follows…

Paying off technical debt

Generative AI can tackle the technical debt that is the inevitable by-product of digital transformation initiatives.

Our recent research found companies, on average, are spending 21% of their IT budgets on managing this expanding technical debt. The same technical talent and resources needed for clearing technical debt are also needed for building digital transformation initiatives that create it. Something of a productivity paradox for businesses. However, generative AI models can automate the process of rewriting outdated code (a common manifestation of technical debt), which often plagues larger organisations with expanding digital infrastructures.

Generative AI’s ability to understand and solve code-related issues also makes it an ideal tool to help software developers to optimise code. By identifying where ChatGPT can contribute to tasks like legacy code conversion or automation of processes, companies can free up valuable developer time for building the tools that can grow the business or drive new initiatives. An added benefit of this use for AI is that developer teams can automate some of the coding of new applications or systems, by updating old code or re-using existing examples.

ChatGPT can turbo-charge development time by scouring swathes of information on GitHub and other forums to find examples of what you’re trying to build. It can do the heavy lifting here, allowing developers to verify, adapt and fine tune the code before launch. Developers will become more productive and businesses will get faster feature development and quicker releases.

We recently released our own API Connector for ChatGPT, inspired by positive industry use cases that we see, especially in process-heavy operations such as the supply chain.

Beyond “Hi Subhash” personalisation

Software AG developers created a ChatGPT API Connector for webMethods, which allows companies to integrate OpenAI’s new tool into their existing business processes to enhance numerous functions, like customer service, contextual feedback, or ticket creation.

SoftwareAG’s Ramachandran: Getting ‘real world practical’ on generative AI for developers.

This connector will allow organisations to quickly start testing ChatGPT’s ability to help with their processes, for example around product or service fulfilment. Plugging ChatGPT into the process tracking around the delivery cycle can enable automated updates to be generated.

AI can create personalised messages – not just “Hi Subhash” personalisation, but an explanation of my specific circumstances, because it’s able to understand very quickly every aspect of my order.

It’s crucial, of course, for all of the right data to be connected and available for this kind of personalisation to be possible. But empowering an AI with this data allows it to understand the situation in a way that automated systems haven’t before.

Humans, augmented

This could open up a much more frequent and informative dialogue with customers at a scale that hasn’t been possible with a human-only process. It’s not an approach that does-away with the human element of customer service, but it is one that allows communication with a much bigger pool of customers, pre-empting complaints or enquiries and protecting that valuable resources of human agents.

For enterprises that are seeking to create the connected experiences their employees, partners, and customers crave, generative AI has great potential, however it’s just one piece of the vast puzzle that is the connected enterprise.

ChatGPT, Bard and other AI can only get smarter if businesses have their internal processes and applications connected to enable machine learning on datasets that are accurate and holistic.

As with any new technology, there’s more to be done by the industry to address the risks associated with Generative AI such as reducing erroneous answers, improving data security and access management, as well as determining who is ultimately responsible for its governance within a business. While businesses are approaching Generative AI with caution, there are many positive opportunities if the technology is integrated properly within a business’ operational processes.

Data Center
Data Management