#WEF19 AI Panel - Learnability is key to the success of Enterprise AI

Such a great opportunity to be on a breakfast panel at #WEF19 #Infosys lounge with Andy Maguire (Group Chief Operating Officer, HSBC), Illah R. Nourbaksh (Professor of Robotics, Carnegie Mellon University) and Kevin Roose (New York Times) to discuss the “Future of AI: Driving Purpose and Profitable Growth”. Some re-assuring comments by Andy, that if we continue to educate everyone coming into the workforce and prepare them for the jobs of the future, it’s all going to be okay! Illah concurs that Technology fluency is becoming an integral part of society and for #AI to be successful, technology and social innovation need to go hand in hand; so much so that now, a #Humanities class to teach the ethical implications of AI is being taught at Carnegie Mellon University.

 I completely agree with the thoughts shared by the panelists as #AI as a topic comes up all too often these days in conversations with clients and the debate is largely around how to create an enterprise wide AI journey, that is Strategic, Purposeful & yet Profitable so that it appeals to the CEO and the board. Analysts estimate that AI will save the banking industry over $1 trillion by 2030, so how does one create a well embraced AI program that gets us there? It is often too difficult, given the dizzying array of AI/ML products and platforms available to be firmly placed on this journey.

We have seen some initial success with #InfosysNia with clients as it allows organizations to straddle different approaches simultaneously and decide how fast or how slow they want to pursue this agenda. Here are some options:

Rapid qualification of use cases, through a AI Garage model, that leverages learnings from hundreds of use cases across industry verticals

Democratization of AI within the enterprise, setting up an AI platform that can be used by anyone within the enterprise allowing organizations to scale their AI innovation

Pre-built Horizontal Solutions and Vertical Apps that can be quickly deployed; some examples are Contract Analysis, Customer Contact Center Modernization, Smart Search, Working Capital Utility, Procurement Insight, Demand Forecasting, Promotions Effectiveness, etc.

 We have had success with pre-built apps that have been quickly deployed, for instance –

• Loan-loss Mitigation App: We’re saving a US regional bank 25 bps in charge-offs with our loan-loss mitigation app, by predicting delinquencies and recommending actions and investments
• Contracts Analysis App: With bio-inspired parallel pathways for vision and text processing, has helped a Japanese hi-tech major liberate and re-invest 1000 contract managers’ time while preventing revenue leakage, and ensuring better compliance and risk mitigation

Besides these we continue to build custom solutions by helping clients choose the right use-cases to maximize AI returns across the front, mid and back office. Some live examples on these are -

Front-end Customer Experience Transformation:

• We’ve helped a US mortgage servicer accelerate and scale loan on-boarding with 50% less effort
• With our chatbot, a leading credit card company can now resolve supplier queries in ~1 second, instead of 3 days

Mid and Back-office use-cases:

• We saved a US financial major $60 M by predicting customer churn and targeting customers for retention.
• A payments services client saved $2.2 M with our 93% accurate fraudulent transactions prediction.
• For a global Top15 bank facing increasing costs associated with growing loan applications, we delivered AI-led end-to-end process automation, reducing timelines by 50%, and operating costs by 60%

Even more exciting are the new ways we’re finding to unleash AI’s potential. For all of this to mature and thrive, I think learnability within the workforce is key. That’s the journey that Infosys is on, so that the extent of change AI can bring to our clients is only limited by our imagination. What are your thoughts and what boundaries are you pushing so that #AI is a strategic reality for your organization?