Talent problem lies in failure to upskill employees

A $320 Billion GDP boost by 2030, according to PwC. That is the potential economic impact of artificial intelligence in the Middle East. The United Arab Emirates can look forward to the largest relative impact as part of this technological boom. A projected 14% of 2030 GDP will be attributable to AI in the Emirates.

Nobody wants to be left out of a boom. AI adoption has been gathering pace for years in the UAE, but in the wake of one of the worst economic downturns on record, companies looking to become AI-powered still face talent gaps. This remain the number-one barrier to effective AI adoption.

UAE firms have notorious difficulty in hiring the right AI talent because it is both in demand and scarce, and therefore prohibitively expensive. But part of the talent problem also lies in a failure to upskill existing employees in AI.

Assuming the will to build an AI culture, here are some dos and don’ts as you embark on your talent hunt.

Do: Write clear job descriptions

Data scientist is not the only role that will make AI a success. AI, like any adopted technology, has a lifecycle, and many roles are needed to make it work. The trick is knowing what you want to do and who you need to make it happen. So, write job descriptions based on your organisation’s needs.

Data scientists, for example, work with data to build algorithms and models that solve business needs. A ready-baked data scientist will be someone with expertise in machine learning and other AI areas, working knowledge of statistics, and a sound business head. But a statistician could be trained to become one if time and budget allowed.

A data engineer could be described as being a more technical resource, responsible for infrastructure and connectivity to data sources. But today’s visual data tools have all but negated the need for them.

Do: Strike a balance

Hire too many data scientists and not enough data architects, and you may, for example, have problems with connecting to data sources. This could lead to a breakdown in the project-delivery pipeline, which in turn could lead to attrition in both talent areas. Also carefully consider the appointment of managers and data leaders. They should be clear-headed not only on technologies but on a range of business and interpersonal matters.

Do: Invest in remote-working tools

While the home-based AI professional is traditionally more productive, their employers need to be ready for the challenges of system access, security, collaboration, and the effective reuse of past successes.

Do: Hire a diverse team

Responsible AI emanates from diversity and inclusion. Restricting data and AI operations to highly specialised, agile teams can lead to quicker times to market, but at the expense of long-term benefits such as scalability, sustainability, and the democratisation of data processes. Collaboration between people of different strengths, backgrounds and educational profiles can identify issues within AI designs that may prevent the finished product from delivering an equitable performance.

Do: Create upskilling paths

Retaining talent means taking responsibility for its development. Upskilling needs to be individualised, while still serving business needs.

Don’t: Wing it

While writing job descriptions, it helps to consider what projects the employee will be delivering. As long as the skills fit the deliverables, you will have taken an important step in building an AI culture, but it is important not to go after a job role data scientist, data architect, analyst in isolation of clear business objectives.

Don’t: Look for unicorns

A highly skilled data scientist with a PhD and an encyclopedic knowledge of all things AI is likely outside your catchment abilities, unless you are Google, Microsoft, Facebook, or a similar titan. The good news is that most of the skills you need to succeed are found in many non-unicorns. A good data scientist is a good communicator, both verbal and visual, through presentation tools both basic and exotic.

They are business-oriented and have a basic understanding of finance. They are excellent with statistics – grasping the difference between correlation and causation and recognising representative datasets from ones that are skewed. And they are comfortable with any scale of data, from the modest to the vast.

Don’t: Neglect your culture

There is a temptation to let technology take over and forget who you are. Always remember that data teams are driven by knowledge, so investment on training is always going to yield dividends, both in terms of productivity and talent retention. Data professionals obsess over efficiency, so investing in productivity tools, and adopting the right methodologies will be useful. In addition, data teams are driven by seeing their work actualised, so take care not to initiate a project that you are not confident will be deployed across the organisation.

Don’t: Neglect existing talent

Many companies do not realise they can already staff for an inclusive and sustainable AI strategy through upskilling. They have within their ranks eligible employees that can be trained to fulfil many data-driven roles but are not.

Addressing this neglect will also address talent retention, because if you show faith in an existing employee and empower them to become a key resource, the likelihood of them looking elsewhere for opportunities is that much less than for an AI professional you hired from a sparse market.

Agile and ready for anything

A regional business that is looking to AI is doing so for one broad purpose. Knowledge is power; and knowledge in a crisis is survival. The lessons of crisis are still ringing in our ears following two years of harsh education by the pandemic. AI paves the way to a more agile and futureproof business model. And the good news is we can have it. The talent is there. All we need to do is find it and keep it.

Key takeaways 

• Hire too many data scientists and not enough data architects, and you may have problems connecting to data sources.
• In the wake of one of the worst economic downturns, companies looking to become AI-powered still face talent gaps.
• A data engineer could be described as being a more technical resource, responsible for infrastructure and connectivity to data sources. But
• Today’s visual data tools have all but negated the need for data engineer.
• Data scientist is not the only role that will make AI a success.
• AI has a lifecycle, and many roles are needed to make it work.
• Retaining talent means taking responsibility for its development.

UAE firms have difficulty in hiring AI talent because it is in demand and scarce, while part of the problem lies in failure to upskill existing employees.