Ready to learn Machine Learning? Browse courses like Machine Learning Foundations: Supervised Learning developed by industry thought leaders and Experfy in Harvard Innovation Lab.
Enterprise AI is the new hot topic in technology, especially as the consumer space blossoms with sales and adoption. However, the assumption cannot be that the same approach for the consumer market can be taken directly to the enterprise. Consumers push the expectations of AI for the business to new heights – and if not carefully prepared, solutions will inevitably fail. In fact, we have seen this already, as software vendors – ranging from start-ups to large software conglomerates – watch users struggle to adopt, understand, and ultimately see the value in AI.
In the consumer space, customers are generally willing to embrace new technology, take more risks, or even understand the current limitations. However, in a business setting, too much is at stake. For instance, a user cannot trust AI solutions to fully execute or even assist with a potentially multi-million-dollar task related to production, sales, or distribution. Users cannot assume the same risk, with so much more on the line. Therefore, the skills and objectives must be clearly defined, with specific parameters, in addition to constant feedback cycles.
This is not an easy task to achieve, especially when it comes to the same issues that enterprise solutions are facing today, such as complex integrations, data aggregation, upgrades, and more. AI software providers will find success by capitalizing and placing an increased focus on the following:
Self-sustaining data repository
Stitching solutions together to keep them in tune is not easy, especially for cloud solutions where IT has limited access and instead relies upon support. As we still exist in an age of modifications, although decreasing, the upgrade process for newer technologies can be a nightmare, specifically for AI.
This can all be made easier if access to the data, and the tools themselves, are part of the solution – from the beginning. We all know that the value of AI is completely driven by the breadth and quality of the data. Layering services on top of a separate repository is not enough. The best solutions will allow for data ingestion and training off a data lake. This will help control the complex integration points to prepare, cleanse, or even update data points. Users should not be responsible for this maintenance.
Embedded security model
Security in the age of enterprise AI could arguably be the sleeping giant. Not every business model can take on additional security risks. Today, with voice-driven solutions, this opens an entirely new form of security threats. Some solutions may have controls in place, but how do you tie that back to your core business security?
Perhaps the answer is to authenticate a user based on voice, PIN, or even facial analysis – but, how do you then effectively acknowledge their role, association, groups, or any other security defined protocol with their business environment? Unless this is built into the core solution, the same stitching problem exists, only this time it pertains to security. Bet on the vendors that can account for this dilemma from the onset.
Constant feedback mechanisms
Feedback cycles are a required checkbox for any AI solution. Without proper audit trails or validation links, enterprise users will never fully rely upon the technology and achieve its true potential. An AI platform must account for visual proof of data and execution, access to validate the source of the information, and even data behind systematic decision-making.
Humans sometimes struggle to describe their own decision-making parameters, and taking this to the next level with deep learning adds even further complexity. Feedback mechanisms should be considered as creatively as possible in every modeling environment.
Citizen developer toolkits
Lastly, AI is still an acquired knowledge. We rely upon expert code and data scientists to define the experience. In the same way that citizen developer terminology became the keyword of software development, data science will follow the same trend. While the mathematics and development skill-set will continue to be a critical need, a new layer of experience must be delivered to capitalize on the opportunity that AI presents. Users should not feel scared to experiment with capabilities due to limited development knowledge.
As technology continues to innovate, consumers use what is currently available to explore their options. They do not necessarily need to know the how or why; they just need to see the results. This is what ultimately drives technological revolutions – building upon knowledge with new critical thinking is made possible by using previous innovation. Spielberg may not know the inner workings of a cinema camera, but he still directs masterful films through how it is used. The same notion should be considered with AI: focus on the new target experience and how users can innovate, not on the technology itself.
