Sanjeev Vohra, the Global Lead of Accenture Applied Intelligence shared with us how AI & ML has taken over production firm. For this, Accenture have been working very closely with the academia to see how they can translate value from data and put to future use.
Given the fact that we are in exciting times which involves the emergence of AI and cloud that has the potential to unlock the true business value, the last decade has been tremendous in terms of value. Though the expectation from the future is huge, but primarily, he believes that there are 3 things that has helped this transition move to the forefront. The first is data itself, which has grown by leaps and bounds. The immense amount of data, which translates to more than exabytes of data being generated everyday by machines, leads us to question what can be done with this intangible wealth. However, 80% of that data is unstructured and consists of things like images, audio files and documents and leaves us to decide how effectively we can capitalize on that.
The second is AI and its associated technology which has grown at an exceptional speed in the last few years. Adoption of these technologies have grown and the pandemic has only accelerated the process. ML models have been put to production at scale. He cited the example of one of his clients in food business who have put more than thousand ML models into production to enable dynamic pricing, loyalty, customer service, supply chain process, etc.
The third and the most important one is the expectation of the customer from business which is leading to the discussion what is the value that AI can add here and how it can be put to use. This expectation has been further fuelled by the digital native companies who have no legacy systems to manage and open towards adoption of latest technologies.
To make the listening more interesting and audience’s understanding clearer on how unstructured data has been dealt with, he cited the example of an environmental foundation in Philippines who teamed up with Accenture to use AI and ML to save their coral reeves. One of the islands of Philippines, which is famous for its biodiversity, started showing erosion of its reeves due to over fishing, under water trawling and tourism. The objective was to find and trap the health of the reeves and then take actions based on them. They identified multiple parameters for the project but the most important of them were to identify the number and variety of fishes which is an indicator for them. The process, involved using non-intrusive ways like IoT, sensors, cameras and using them with minimum disruption in a fragile environment to gain understanding. Interestingly, backed by computer vision and deep learning, over a period of time, the team was successful in creating 70000 images and audio files and use them to find out the variety of fishes. The data, thus produced, was also used for tracking the remedy actions and the progress achieved.
He then touched on an important point on how to collect and qualify for the right data sets. To understand this better, Accenture carried out some surveys and interviews and based on the findings there are 3 important points that stand out amongst others. First of all, this topic itself has become a C-suite topic which was not the case a few years back. In fact, 60% of the CEOs of the world’s largest companies have mentioned AI publicly. This implies that “they are looking at data as a corporate asset.” Secondly, everybody understands the need for investing in building a platform which can source and curate data sets real time and can generate insights and take action. This calls for the adoption of clouds. The third trend that is emerging is appreciation of the fact that talent is required which, in itself, is scarce. Hence, there is a need to build talent that can build and maintain these models. He signed off by saying that “it is not about the big data but about the right data sets, which could be hybrid or a blend of structured and unstructured, and qualifying them the most critical and hence so close to business discussion.”
Source: indiaai.gov.in
