Virtual QA Roundtable: Driving QA in an AI-Led World
Infosys Quality Engineering hosted a Virtual QA Roundtable on ‘Driving Quality Assurance in an AI-led world’ on May 21, 2020.
As businesses adjust to new normal in wake of Covid-19, technology advancements and pace of adoption will be accelerated to create more resilient organizations of future. Artificial Intelligence (AI) and Machine Learning (ML) are at the core of IT and business strategy for Digital Enterprises aiming to deliver personalized customer and employee experience, reduce opex, prevent risk materialization and create a resilient and Live Enterprise. Quality assurance of AI systems has been a territory less explored and needs formal and structured approach to ensure systems are robust and safe.
This roundtable focused on preparing industries for how deep learning solutions can be leveraged to change business outcomes, in collaboration with academia. It was well received by the audience and attendees showed their interest by sharing their opinions and asking pertinent questions in the area of AI.
Keynote Session – Building a robust and safe Digital enterprise in an AI-led world
Dr. Jim Whitehead, Professor of Computational Media, University of California, Santa Cruz
Dr. Jim’s talk was centered on AI lifecycle in general and QA challenges for AI in particular. He also touched upon Explainable AI, which is visualization approach for understanding the behavior of an AI system. In summary, understanding why models make errors is key to fixing them, thereby necessitating the need for AI/ML-based QA solutions to address them.
Expert Session – Explainable AI - Applying formal methods to analyze and verify neural networks
Dr. Corina Pasareanu, Technical Professional Leader – Data Science @ KBR, Ex-Associate Research Professor, CyLab, Carnegie Mellon University and ACM Distinguished Scientist, Robust Software Engineering (RSE) group, NASA Ames Research Center
Divya Gopinath, Researcher in Formal Verification, Robust Software Engineering (RSE) group, NASA Ames Research Center
Dr. Pasareanu and Divya discussed broader contours of Artificial Neural Networks, their applications such as Pattern analysis, Image classification, Sentiment analysis, Speech/audio recognition and Perception modules in self-driving cars. They emphasized on related challenges which can be solved to a large extent, by formal verification methods using AI/ML capabilities.
Session – Leveraging Deep Learning to Change Business Outcomes
This session was by Rajeshwari Ganesan, Associate Vice President, Infosys NIA (Purposeful AI Platform). She shared insights on real-world challenges in QA of AI/ML Systems by analysing capabilities of a product, its business use cases, practical challenges associated and how they can be reduced significantly by integrating AI/ML in QA solutions.The final Session was delivered by Harleen Bedi, Senior Industry Principal, Infosys Quality Engineering. The talk was focused on QA evolution and how AI/ML will be inseparable part of next-generation QA tools. There is a need for different types of QA strategies at every stage of development cycle, starting from writing user stories to intelligent automation with AI to their final execution. AI/ML will be integral to fool-proof execution of continuous DevOps projects in a speedy and timely manner. Infosys AI/ML platform for testing has the ability to support development and testing of newer ML use cases in future, improving QA effectiveness at every level of QA lifecycle.
Below is the event’s complete recording for your viewing.
To know more about our AI in software testing offering click here.