The importance of AI has never been more widespread, with every industry, every business discussing its merits at a boardroom level. To be clear, AI will bring fundamental benefits to the organisation, but as with every new technology, there are complex challenges that lie ahead if anyone seeks to realise its potential.

In the previous posts, our experts discussed how the state of an organisation’s IT infrastructure and data storage system affects the value it can derive from AI, as well as the security concerns that accompany the use of AI.

This is issue #3 in a series of four articles on the expert perspective on AI readiness, and here our thought leaders discuss how an organisation’s talent and culture contributes greatly to the success of AI implementation, and whether it can get the best value from it. Here are their top tips:

  1. When hiring talent for AI, what skillsets do I need to ensure the smooth implementation and operation of AI?

Rohit Pandharkar, Mahindra Group

The candidate should have a good grasp of programming languages such as Python and R, and by extension, should be adept at writing code. They should also understand machine learning algorithms and prediction cases – a demonstrated ability to build algorithms is crucial.

Building AI naturally involves a lot of testing, which involves applying algorithms on real systems and making constant adjustments on parameters to get to the eventual application of the concept. Candidates who are able to code the algorithm will find more success with this whole process.

  1. What can my organisation do to attract the right talent to implement AI successfully?

BS Teh, Seagate

Getting the right people for the job is vital for a complex task like AI implementation. According to the Data Pulse findings, almost all (94%) respondents currently feel their organisations could do more with AI if they had the right talent. More than that, one-quarter don’t believe the best talents have been assigned to develop and implement AI.

To attract great talent to establish an AI platform that drives your business forward, start by building the right organisational culture. For AI to thrive, build a data-driven organisation. This means creating an environment where data is seen as the organisation’s metrics for success – data needs to be easily accessible, with the right tools provided to easily derive actionable insights.

In the face of rapid change, the organisation also needs to commit to reskilling and upskilling existing talent. Demonstrate an inclination to nurture the next generation of data specialists – management trainee programs and internships are highly effective tools to build the organisation’s talent pipeline while bringing in diverse perspectives to approach challenges in new ways.

Finally, it’s important to establish a culture that embraces AI. We need to change the borderline ‘apocalyptic’ view that a world with AI means the death of all jobs. AI will enhance human capabilities and bring benefits to the organisation and employees, and this should be embraced across all levels of business.

Akash Bhatia, Infinite Analytics

As with any hire it’s important that the hiring manager knows exactly what they are looking for, and to ensure it is clearly communicated – both externally and within the organisation.

AI is a very cool term these days so you can expect all kinds of people will flock to any openings. If your organisation does not know what skills to look for, it could be forced to review a huge number of applicants unnecessarily. Hiring a consultant to identify those skills would be beneficial in that situation.

This increases your chances of better hires in the future – your new hire will create the right conditions and assist in setting the benchmarks of who and what to look for.

Johnny Chou, Viscovery

With all the current conversations surrounding AI, it’s important for businesses to avoid jumping onto the AI ‘bandwagon’ too blindly. An organisation must develop a holistic view of its need for AI, and understand fully how its business can benefit from AI implementation.

This is especially pertinent in the area of human talent. Although deploying AI may streamline processes and reduce the demand for manpower, AI technologies, for the most part, are meant to co-exist with human employees.

Human talent is an important component in collecting and analysing data that trains AI platforms – important for keeping pace with the ever-changing business environment.

Therefore, organisations must pay more attention in cultivating their talents, and encourage employees to continuously iterate and improve AI work processes, which will ultimately improve the overall workflow.

Rohit Pandharkar, Mahindra Group

To attract the best AI talent, it’s important for the organisation to establish itself as a thought leader in the space. To do so, it could host or be involved in data science meet-ups to be present in the minds of potential talent. It could also consider hosting hackathons, presenting challenges in multiple areas of machine learning to showcase the interesting work that happens on the inside. Alternatively, it could build its profile by showcasing best-in-class computational architecture so that potential talent will be attracted to the prospect of working with the latest innovations in the industry.

  1. What kind of training is required to train employees to use AI effectively?

Akash Bhatia, Infinite Analytics

A strong foundation in math, statistics, and programming is needed to be successful with AI – your training needs to help your employees build an expertise in these areas. However, it’s important to note that merely investing in a three-month course in data analytics and machine learning isn’t enough. This is a space that continues to evolve and grow, and the organisation needs to continuously invest in AI training to ensure employees are always on the pulse of what’s new in the space.

Rohit Pandharkar, Mahindra Group

One important area to get employees up to speed is to train them in using deep learning infrastructure such as graphics processing units (GPUs). They also need to know how to measure the impact of the algorithms deployed, and the accuracy of the AI implementations undertaken. More importantly, employees need to learn how to translate business problems into data science problems so they are able to identify areas of the business that can be improved, and how to implement AI to solve them.

  1. Between hiring experienced talent with a full set of capabilities and training more junior staff, what is the best talent management strategy to implement AI effectively?

BS Teh, Seagate

To implement AI effectively within an organization, I believe the answer to hiring talent lies somewhere in the middle. Hiring a mix of experienced and less experienced professionals mean that there is more room for the organization to grow. Senior professionals can help with mentorship, and junior talent will come in with fresh eyes and innovative ways of solving problems in this rapidly changing environment.

Additionally, don’t forget about your in-house talent. Helping in-house talents upskill alongside existing resources who know your business reduces the dependency on hiring full AI experts, who are harder to come by.

Akash Bhatia, Infinite Analytics

For anyone building an AI roadmap, start with hiring experienced people with a full set of capabilities. Given that AI is a relatively new area, these openings will take a long time to fill.

While you search for the superstars of AI, you should also upskill junior talent within the company – it’s another long-term project that largely depends on the urgency of your implementation.

In my opinion, the best talent strategy is to bring in experienced people along with a few graduates from math and statistical institutes across India.

Johnny Chou, Viscovery

Organisations should first consider the nature of their challenge, and if deploying AI in this instance would really solve it. Urgently hiring a group of AI experts without understanding the real challenge of their business would undoubtedly end up being a waste. Furthermore, training in-house employees with little understanding of the business or those who are less qualified would also be a useless attempt.

Ideally, the management team needs to understand AI technologies at a certain level as well as its benefits and risks. Aside from training representatives from various departments, these managers should develop a mid-to-long-term plan to introduce AI into its operations and prioritise the business areas for deployment.

If an organisation lacks the requisite AI knowledge in its initial stages of implementation, it could also seek assistance from AI consultancy firms.

Rohit Pandharkar, Mahindra Group

An organisation cannot be successful with an either/or scenario in this case – I believe a healthy balance of both would be required. In fact, I like to think of 1:3 as an idea ratio of experienced staff and juniors. For example, there should be one experienced team leader/ program manager who runs a team of three to four data scientists, where the senior leader takes on the responsibility of identifying and translating business problems into data problems, while training the rest of the team to do the same.





Leave A Comment