Unicorns are a rare breed not only in the world of startups, but also in AI talent acquisition.
A candidate described as a ‘unicorn’ has specialist skills, niche knowledge and experience, exactly the right qualities required for the job, and can communicate effectively to business users.
In short, they’re a perfect match!
As AI adoption continues to grow and business value is realised, organisations will aim to increase their talent pool by recruiting data scientists and specialists that can do it all.
But when it comes to scaling AI, the top two blockers for companies are hiring people with specific skills and identifying good business cases.
With expectations of candidates to address both issues, it’s no wonder that data scientists can often feel like unicorns.
The increasing prevalence of AI is shifting the focus towards delivering value-driven outcomes, fueling a need for workers with tech-specific skill sets.
However, hiring more data scientists doesn’t quite solve the dilemma of scale, because they’re in high demand, difficult to attract, and even harder to retain as they become more experienced.
Outsourcing is a possibility, but according to McKinsey, is no longer your way out of talent problems as core capabilities need to remain in-house.
Additionally, the ambitious goal of having 1.2 million tech workers by 2030, doesn’t quite tally with the current skills shortage plaguing the domestic labour market, making it even more challenging than ever before to hire data and AI experts.
What’s increasingly evident is that hiring specialists, unicorns no less, is no longer the answer to building an AI-driven organisation, particularly at a time when budgets and resources are stretched.
Instead, businesses need to think about moving AI mainstream, by tapping into collective intelligence and building communities of interdisciplinary business and data professionals across the entire organisation.
With the drastic volume of data generated daily, business analysts need to move on from spreadsheets and support business decisions with data-driven insights; likewise, data analysts need to understand and communicate the wider business implications of data.
To bridge the talent gap, establishing hybrid teams that encompass business intelligence and data intelligence is the way forward.
Here are three proven strategies for success:
1. Adopt a common AI/ML platform
A prevalent issue with scaling AI is that innovation is only in the hands of a few experts.
Many AI functions developed today are fairly similar to how things were made in the pre-industrial revolution era, produced by artisanal experts in small batches, keeping knowledge closely guarded.
This means that productivity is low but maintenance and costs are high.
Similarly, businesses struggle with the complexities of bringing AI to fruition with expertise being in the hands of only a select group of business specialists and analysts. When other members of the organisation are unable to understand AI models applied to projects, business continuity and shared institutional knowledge is at risk.
What organisations need to strive for is shifting AI from being something that only a small group of people can access and understand, to an Everyday AI model that permeates across the business and all functions.
Weaving data into an organisation’s DNA is necessary for their survival and that’s where a common AI/ML platform helps unite workers of all skill levels to use it, without eliminating the ability for data scientists to continue developing models.
The optimal platform would facilitate a distinctive collaborative space that brings together AI consumers and AI creators, regardless of their coding abilities.
2. Invest in an adoption program
Any technology investment is only as successful as its adoption; it doesn’t matter how great the platform may be, if nobody uses it, then no value is generated.
Likewise, data scientists using the platform may be creating excellent models, which aren’t going to contribute to output if not used in business workflows.
The goal of an adoption program is to evangelise the use and application of the AI platform to move it from a stage of intent to action, both from a data worker and business workflow point of view.
This can be done through several activities such as training programs, communication campaigns, and regular monitoring of satisfaction metrics.
You could even inject a bit of fun and strengthen awareness through branded badges and laptop stickers or run hackathons and internal competitions to celebrate wins.
3. Create an upskilling program
A successful upskilling program creates a positive cycle where business analysts learn AI capabilities and add value.
This increased awareness of AI and knowledge capacity invariably leads to identifying new users and a new cohort of data workers looking to be upskilled.
Whereas an adoption program is designed to encourage the rapid uptake of a new platform, an upskilling program ensures the benefits are widely spread across the organisation through tools such as self-service training.
Supported by a common AI/ML platform, there are two models of upskilling programs that have been successful in our experience: interdisciplinary (grouping workers with the same skills regardless of the business function they’re from) and functional (workers with similar skills in the same business functions learn together).
Regardless of which model works best for your organisation, the long-term goal is to create a sustainable community of professionals and drive ROI across many business units and functions.
Closing the AI skills gap
The core of an enterprise’s data strategy should augment AI capabilities with human capital, making data science less mythical and more achievable, less restrictive and more accessible.
Data scientists are consistently in high demand, resulting in higher tendencies to switch jobs, so businesses must build a culture that amalgamates data with its business functions.
Scaling and employing a robust data methodology is pivotal to becoming a true AI-driven enterprise.
To do so successfully and sustainably, enterprises should look beyond hiring data unicorns and build unicorn teams to close the AI skills gap.
Uniting people from across the organisation regardless of skill level and function, will not only help gain broader insights from their data, but also enable them to transfer their knowledge, work faster and smarter together.
Doug Bryan is Field Chief Data Officer at Dataiku.
This content has been written by a topic area expert and is not a sponsored post or advertisement.
Information Age welcomes Opinion pieces from industry leaders. You can find our submission guidelines here.