Alan Jacobson, Chief Data and Analytics Officer at Alteryx, discusses the importance of upskilling employees to keep up with rapid advances in AI technology.

Many of the world’s leading companies recognise the vital role that digital transformation plays in their success. Despite this, however, their pace of transformation can often be held back by outdated code-heavy technologies and processes that make it challenging to unlock value from data.

With data types and volumes continually increasing, the challenge is figuring out how to accelerate time to insight and make decisions from this sea of data. A large part of the problem is that many organisations have been see-sawing between hiring for specific coding and programming skills to try and leverage business data without considering what’s really needed to achieve their ambitions of becoming data-driven.

Multi-skilled workforce

While there will always be demand for skilled developers and data scientists to code more sophisticated applications, the appetite and capability of business users to create workflows that solve domain-specific challenges is growing. For instance, AI and machine learning technologies offer tremendous opportunities to deliver data intelligence at scale. With proper application, generative AI enriches the data interaction landscape by adding a robust, intuitive layer of engagement that empowers those with no data science skills to deliver insights via a natural language prompt.

However, for its true potential to be realised, businesses must ensure users can access high-quality datasets. To do so, data pipelines must be created by governed analytics processes and developed by organisational domain experts who understand the shape of data and have been upskilled to become empowered with generative AI capabilities. However, many make the mistake of adopting modern analytics tools without considering the range of technical and non-technical employees using them.  

Of course, coders, programmers, and system architects will remain essential for businesses moving forward to code more complex solutions and to monitor the thousands of applications and algorithms that help them to make their digital decisions. But generative AI-augmented, self-service, no-and low-code platforms allow organisations to drive ‘citizen development’ through increasing data literacy – making it possible for people with functional domain knowledge to contribute to an organisation’s analytics and AI innovation.

For example, a code-friendly data platform means a business analyst with no coding skills can work alongside a data scientist using Python or R in the same workflow or model. With 82% of global IT and business decision-makers seeing AI as impacting what their organisation can achieve, there’s clearly a need to upskill the entire workforce rather than depend solely on talent acquisition.

Companies that take advantage of self-service technologies that blend generative AI capabilities into the analytics experience expand the reach of AI and machine learning innovations to wider teams rather than restricting them exclusively to the domain of data scientists. The result is a better experience for non-technical users. This optimised, accessible modern experience to analytics and AI provides an environment where decision-makers, analysts, data scientists, and developers can ‘speak data’ and collaborate to deliver insight through automation.

Bespoke programmes

This approach to AI upskilling doesn’t mean all employees must become data scientists. Instead, it means ‘citizen developers’ are capable of ‘mass-producing’ simple applications that answer questions and deliver insights at the speed the business requires rather than the development team can provide. To do so, leaders must build a multi-skilled workforce that feels confident and empowered to use data and AI in their everyday roles while retaining critical thinking and ‘human’ qualities.

In practice, this means upskilling employees — the accountant, supply chain analyst, and merchandising analyst — so they can leverage their domain knowledge and interest in harnessing technology to incorporate generative AI for core business problem-solving. The value of lowering the barrier of entry, which generative AI enables, cannot be understated. When paired with domain-specific business contexts, transferable soft skills can be a sound basis for using these applications and are vital for enabling large language models (LLMs) to deliver the most effective outputs.

The art of extracting valuable insights from enterprise AI applications doesn’t depend on sheer technical skills alone. On the contrary, it’s easier to train an accountant to use AI-powered analytics in their everyday job, for example, than to train a data scientist on the nuances and functions of accountancy.

Evolving training

With organisations at a pivotal point in the journey to harness generative AI for insights that deliver business value, the tipping point to success is a data culture developed for the era of intelligence. While this journey may initially seem overwhelming, good-quality data is the lifeblood of AI, so it’s critical to have data-ready humans armed with the business context to the questions and the skills to train generative AI to solve their business problems. To do so requires continuous data literacy upskilling so teams that know the business understand the data pipeline lineage and know what it takes to pull the right predictive and prescriptive insights.

Therefore, a good starting point for an AI upskilling programme is an assessment of current technical and soft skills within a workforce, around which a bespoke training programme for individual employees can be built. Many organisations are turning to on-demand curricula rather than sending employees ‘back to school’. This is an excellent approach to keeping up with fast changes in AI and is possible thanks to low-cost, high-quality content being produced within the analytics community.

AI technology is constantly developing, so any approach to upskilling must be flexible and dynamic. To this end, the best way for organisations to create an internal data-literate culture is to provide accessible platforms that democratise data access and pair these with wide-reaching skills training.