One of the political leaders in India regularly publishes videos in which he spends time with people who actually form the spine of the Indian economy (or the economy of any developing country). These are workers who work in the field, drivers, mechanics, laborers, factory workers, farmers, etc. I love to watch these videos to understand the key challenges that these people face in their daily work and lives.

In one of the videos, this politician traveled overnight in a truck from Delhi to a nearby state, conversing with the driver and his associate. A few months later, when this politician was in the U.S., he decided to do the same with a driver of Indian origin. Obviously, he could not help but notice how nice trucks in the U.S. were compared with those in India. But to me, the crux of the entire video was just one sentence he said during the episode: ” The trucks here are designed around the needs of the driver whereas, in India, drivers have to adjust to the design of the truck.“

Oh boy—this applies to so many products around us, and software products are no exception.

The good news is that software product companies are increasingly using AI to change this narrative.

AI can bring flexibility into software products, making them more “customizable” for each individual user. Many large enterprise software companies are already on the path to building this. But there is another area that is a major bottleneck and can use the power of AI—implementation and configuration of these systems. I touched upon this in yesterday’s episode of “Friday Fun.”

If you conduct an unbiased and maybe confidential survey of the bottlenecks that users of large enterprise systems have faced, implementation and configuration will emerge as a significant challenge.

While the intense nature of implementation was “unavoidable” a couple of decades ago, that is no longer the case. In fact, in less than a decade, we will see a whole new breed of enterprise systems (I will touch upon it in my upcoming report) that can be leveraged within days and will also be extremely customizable. Hence, large enterprise software companies must eliminate this “complex implementation” bottleneck or reduce it to a bare minimum. AI can help!

The exact ways AI can help can be explained in detail for each system and hence will be as detailed as the implementation documentation of these systems, spanning into hundreds of pages. However, as touched upon in the Friday Fun episode, the gist of all the details will be the ability to learn from the existing data, not only to suggest the best configuration but also to perform a significant percentage of configuration.

There are plenty of stories of implementations gone horribly wrong due to wrong configurations during implementations that fail to capture real-world business processes. The cost associated with these failures is mammoth. AI can help guide and accelerate this process, cutting the implementation time and cost by more than half.

This will also significantly enhance the revenue of the enterprise software company. Believe me, if you can successfully build this “Implementation AI,” organizations, who almost always despise the complexity of implementation, will willingly pay you fat dollars if your AI does its job. This reduced bottleneck will also make your product more attractive, thereby increasing sales in general. This is not an option but a necessary capability you need to build since a new breed of systems will be in the market within a decade. To avoid disruption, be the disruptor.