Alan Jacobson, Chief Data and Analytics Officer at Alteryx, offers advice for IT leaders on modernising data stacks for generative AI success.
In a relatively short amount of time, an enterprise consensus has emerged behind generative AI (gen AI). Earlier this year, we found that 82% of business leaders agree AI is significantly impacting organisational goals and almost half of board members are prioritising the technology over everything else. Luckily for them, consumers/employees are on board too. We also found that 79% of consumers feel positively about gen AI.
So far so good – but ensuring the success of gen AI in the enterprise is by no means guaranteed. As such, many are undergoing ‘data modernisation – over half told us as much in a recent survey. But success frequently lies in the use cases that are selected, not in how new the data stack that houses the project. And creating the right gen AI use cases requires that knowledge workers and IT leaders understand the technology well and where the best places to apply the techniques sit. Unfortunately, if you pick the wrong use cases, you will not see the return on investment (ROI) and problems will rapidly ensue.
Modernising the data stack is important, but must be balanced
As organisations build out their data stack, it is important to keep an eye on ROI.
Building a data lake takes significant time and resources (e.g. data engineers) and will cost significant money. And, unfortunately, the act of building a data lake by itself will not deliver significant ROI.
ROI will come when applications, automation and analytics are delivered. These other types of technologies will take on two forms: centralised teams building solutions and democratised teams leveraging analytics and automation. In the former case, this also requires leaders to invest in people to centrally build the solutions. This typically again takes significant investment and tends to focus on larger problems that have good ROI, but take a while to deliver.
Creating AI-ready data shouldn’t mean that organisations overlook what’s going to drive value from generative AI for them – practical use cases. This means honing in on what Large Language Models (LLMs) are good at delivering today, and where they fall short. For example, they’re very effective at summarising mountains of news data. That could form the basis of a myriad of winning use cases.
It’s not just the sources of data that determine the shape of data stacks and their modernisation. Broader macro factors in the business are hugely influential.
In addition to data sources, we’ve found that existing IT infrastructure and technical expertise are cited as the two main drivers of the data stack structure. This helps explain the appeal of hybrid data stacks in the AI age – they’re best compatible with existing IT infrastructure.
Setting teams up for AI success
In carrying out data stack modernisation, IT and data leads shouldn’t overlook how data teams are managed and organised, or the operational procedures. All these factors need to be optimised for a business to successfully adopt and adapt to gen AI.
Ownership of data needs to be crystal clear across the organisation to prevent confusion around where overall oversight of data access and management lies. In a similar vein, enterprises should consider breaking down siloes with a centralised data or analytics function to maintain data as a shared organisational resource. This will accelerate the time to scale gen AI models in enterprises and, therefore, the ROI they ultimately bring.
Employee skill sets also need to be assessed. As already mentioned, the promise of enterprise gen AI is in its ability to lower the barriers of entry to work with data and generate first-class insight from it. Non-technical employees need to ‘speak data’ and be data-skilled. This could require some level of upskilling.
This audit of broader business process could also unearth some difficult but important questions regarding IT spend. It’s often the case that IT budgets aren’t reviewed or adjusted throughout the year, even if new needs arise. Whether that approach still stands up given breakneck speeds advances in ge nAI remains to be seen. Enterprises that stick their neck out and invest in this technology in quicker cycles will be the first to feel its benefits.
Key takeaway
Delivering data stack modernisation while driving tangible ROI is critical to the success of your analytic efforts. The business must see wins along the journey to keep focused on and accelerate changing the business to become data driven. This change management process is key to succeeding with gen AI as with all analytics, and the opportunity that lies ahead is well worth the effort.
