When enterprises adopt data, cloud and AI, they must also empower employees to work with the technology. They should educate them on the pitfalls of data sprawl, sensitize them to ethical considerations, and train them to be problem finders, says Balaji Ramanujam, head of architecture, Data & Analytics, Infosys.Â
Data, Cloud, AI: Good on Their Own, Great Together
Companies clearly understand the direct correlation between business performance and powering the enterprise with strong data-driven insights and foresight engine. And yet, how confident are they about unlocking value from their own customer and operational data? Recent researchOpens a new window points us to interesting findings: About a fourth of enterprise respondents describe their ability to unlock value from data and AI as very developed. But that proportion jumps to 57% for enterprises with more than 80% of their business functions in the cloud. This is fully consistent with what we see within our client landscapes too.
Source: Infosys Cloud Radar 2021
Pressured by market and economic forces and buoyed by the gains, enterprises are ramping up investments in the cloud as they grow their investments in data analytics and AI. From experience, they have learned the mechanics of successful implementation â€“ which solutions to use, how to bring together public, private and hybrid clouds, whom to partner with and which data sources are most useful to leverage. Less thought is given to how this might influence employees and how they, in turn, can impact the data programs and their outcomes, either out of oversight or an assumption that everyone will automatically learn to adapt to being part of a data-first organization. While there is no doubt that there is ample potential for data and AI on the cloud to amplify human capability, it is fully realized only when there is adequate human empowerment and nurturing of a data-first mindset.Â
Here are three ways of ensuring that:
1. Alert Employees to the Dangers of SprawlÂ
Enterprises can unlock maximum value from data by creating an enterprise-wide data fabric that provides high-quality, relevant insights on demand whenever someone needs to make a decision. As various functions, departments and entities within organizations produce more and more data, with increasing autonomy and speed, the resulting data sprawl makes it hard to manage the information across disparate repositories and make accurate analyses.Â
Data sprawl can diminish the value of data or corrupt it all together, besides making it hard to control and secure. Combined with cloud sprawl â€“ rampant growth of cloud instances, vendors and applications without visibility and control â€“ it can create a governance nightmare that outweighs any targeted benefits.Â
Companies like Kraft Heinz that we partner with understand this only too well. Their vision for their data on cloud architecture has always been one single data hub for the company that powers both the new digital transformation initiatives and the day-to-day analytics that run the company.Â
Data-smart companies like these ensure that their employees are educated about maintaining data and cloud hygiene, so that they can exploit their data fabric to the fullest. This includes cautioning them against downloading data on their personal devices to beat firewalls, keeping multiple copies of information or uploading company data to an unsecured cloud location.Â
2. Sensitize Them to the Ethics of AI and Data
Widespread cloud usage and data resources provide the perfect platform to scale AI systems and democratize adoption throughout the many functions of an enterprise. But this can, if not governed well, pose significant problems on the ethical front. In fact, managing AI’s ethical implications is considered important enough to merit the attention of policymakers and regulators too.Â
The data harvesting and the machine learning models that leverage the data need to be unbiased. These models, built using fairly obtained data, also need to be explainable when it comes to outlining how the outcome is arrived at. For example, being able to rationally and transparently explain why the model rejected a job application from applicant A but approved the candidature from applicant B.Â
That’s why it is critical to sensitize employees, both AI team members and ordinary users, to this issue and impress upon them the importance of working within the boundaries of regulatory compliance. Employees must also be supported by a good governance framework of strong and auditable risk management practices across AI development, validation and monitoring so that the AI systems they build and work with are unbiased and accountable.
3. Train Them To be Problem Finders
AI teams are usually well-versed in the technical aspects of their job, such as writing algorithms, but are less knowledgeable about the business context. This restricts their contribution to coding a solution for a problem that someone else has to identify, define and break down for them. As the people who understand and work closest with data, AI workers are the best people to discover unframed and unarticulated problems from it, if only they knew how.Â
They need to be taught how to dive deep into data to make unexpected connections and discover new insights. Further, training in Design Thinking and customer empathy can teach them to think like problem finders, along with being problem solvers. In fact, institutions like Posti, the famed Finnish postal service that has embarked on its transformation journey with Infosys Living Labs amplifying its capabilities in innovation, are taking advantage of Design Thinking tactics to nurture their teams of problem finders.
And the Data Culture MattersÂ
While the tools and governance are helpful enablers, empowering people to make data, insights and foresight a pervasive part of how they work requires something more purposeful. It requires a culture that encourages a deep cognizance of data’s growing relevance for business and the need to be curious and exploratory but also cautious and conscientious.