While a majority of AI projects still don’t reach production, the interest for no-code AI platforms keeps rising. Indeed, a growing number of startups and large tech firms now propose “easy-to-use” ML platforms. The idea of being able to build and use a solution based on Machine Learning without being a data scientist is something very interesting for both small and large companies who could empower their employees while dedicating more resources to complex ML projects. In this article, I will share what have I learned after having implemented one of these no-code AI solutions and analyzed several startups related to this industry. As an AI consultant, my goal was to determine if these solutions could help us increase the chance of having more projects transitioning from proof of concepts (PoCs) to scalable, relevant, and efficient deployed AI solutions.
From an operational perspective, we develop several AI projects during the year for several departments. Most of them only remain PoCs due to a lack of data, investments, leadership, or simply due to the current maturity of Machine Learning. When it comes to projects, the level of complexity is never the same. Indeed, some projects are related to “recent” algorithms while others are simply using well-know algorithms such as linear regression, K-means, Naïve Bayes, etc. Our goal with no-code AI platforms was to determine if these solutions can help us transform some of our collaborators into citizen developers and dispatch more data scientists to “complex” use cases.
It is key to understand that the promotion of internal citizen developers such as product managers, doesn’t mean the demand for data scientists is going away. The idea is to lessen the burden on data scientists backlogs so that they can maintain focus on larger, more complicated projects. Obviously, it would be appealing to have to do “less” work regarding the traditional ML workflow and accelerate the development of specific use cases. Would it be possible to envision a future in which managers can come with an idea, build it, and run the project by using a no-code AI platform?
As you can see, some steps of a “traditional” AI project can be automated or made easy using drag and drop tools. From this perspective, I would consider no-code AI platforms as a time-efficient way to speed up prototypes and demo development. Use casesI believe no-code AI platforms are well-suited for specific projects. For instance, use cases in which we want to predict metrics like churn, customer lifetime value, dynamic pricing or analyze data across several contracts to help us better negotiate. We also believe that these tools can be useful in the automation of some internal processes. New skills It is safe to assume that the rise of no-code AI platforms will also create new skills expectations. In the near future, I would not be surprised if a product manager has to be familiar with at least one no-code AI tool and be knowledgeable about dataset management. I expect to see more and more online training related to these tools.
I often wonder if no-code AI startups can survive over the long term if they are not specialized. Indeed, the advantage of large tech firms is the ability to provide customers with a no-brainer approach to staying on that vendor’s platform and roadmap. Ideally, large tech firms (Google & Microsoft) want companies to be able to use their software development tools and a wider ecosystem of services related to data management. While some of these solutions are free or based on a subscription model, they might require the intervention of consultants and developers to train users and perform the cloud back end connection engineering.
For a couple of months, we have decided to test the effectiveness of a no-code AI platform. In my opinion, the efficiency and usefulness of No-code AI are not a myth.
