Leading Machine Learning and Artificial Intelligence Cultural Change
Kaushik Mitra, Chief Data Officer and Head - Actuarial, Data Science, BIR, P&C at AXA Business Services was a speaker at a panel discussion on leading the cultural change for Machine Learning and Artificial Intelligence at the recent 5 th NASSCOM Big Data & Analytics Summit.
In conversation with Kaushik Mitra:
What are the key drivers that influence the adoption of Machine Learning and Artificial Intelligence?
"One of the important changes an organisation needs to action as they look at adopting Machine Learning (ML) and Artificial Intelligence is to look at becoming a data-driven organisation. This mean organisations must begin to leverage the data they have in a more active manner. It is less about the tools used to dissect data and more about the processes established to help all in the organisation leverage data at all levels of decision making. There is a myth that data is needed for complex decisions only, this is not true. Even the smallest decision can be made far more efficient when backed by data. This is when data-driven decision making becomes a culture in an organisation and builds an appetite for advanced Machine Learning and Artificial Intelligence. The transition to these advanced streams then becomes natural and smooth.
What are the trends in Machine Learning?
The industry will make more inroads into Machine Learning powered Artificial Intelligence that will deliver systems that understand, learn, predict and adapt and potentially can operate autonomously. The Algorithm powered economy will become a norm opening up Self-Service BI. Advanced algorithms like neural nets and greedy algorithms will be leveraged further. For organisations, one advice would be to use these technologies to experiment in small clusters and measure incremental wins. Organisations need to focus on becoming data empowered. This means getting the right data on demand at a click of a button - that would mean moving away from how we conduct our board room discussions through standard presentations to leveraging “Data rooms" of the future where data from different silos are brought in to deep dive for any executive decisions. However, while Machine Learning and AI can power board rooms, a completely autonomous organisation is still a very futuristic outlook. Another area is messy data. Machine Learning can help streamline messy data, and in our industry this could mean providing insurance to areas/situations not possible before.
What are some of the new test cases for Machine Learning and AI?
Machine Learning is helping the insurance industry move from risk indemnification to risk prevention. In health insurance for instance, using connected technologies and analytics, users can monitor and be alerted on their health conditions helping them prevent any health issues. Telematics for instance in automobiles help reduce road accidents. We are also able to leverage Machine Learning and AI to introduce new solutions in the insurance space. For instance, in AXA we are looking at providing insurance to Drone pilots. This is a new area and opportunity. Here our labs are leveraging Machine Learning and AI technologies to look at scanning of geo map and the current weather to fine tune the insurance policy and coverage real-time.