The AI Accelerator Festival is running from March 16-19 and we have a few pre-event Q&A sessions to get the ball rolling.
Joining us today is Mahasa Zahirnia, Chief Engineer/Cyber Engineer/Artificial Intelligence Lead at Northrop Grumman.
Her talk is on March 19th - Building a cyber security agent utilizing AI within the product
We caught up with Mahasa to discuss her upcoming talk, what excites her about AI, the biggest challenges for organizations adopting AI and the best projects she's worked on.
So, want a sneak peak into her presentation? Keep reading!
Q: To kick things off, it would be great to learn a bit more about your background. What led you to get into your chosen area of AI?
A: I am AI Technical Lead at Northrop Grumman and work with various sectors to resolve issues that can’t be defined with common engineering solutions. AI allows us to create an actionable solution to various data challenges such as data mining, statistical learning and development of smart technologies.
Q: What excites you the most about AI within your respected industry? What advancements do you see disrupting/having an impact on the areas you work in?
A: I expect AI applications such as autonomous systems, automation, cyber and process improvements will unlock significant economic value, and however the essential key parameters have to be available to the architect for the expected results. They are availability of data, strong algorithms and high computing power.
Q: What is the biggest challenge within your respected industry for organizations adopting AI?
A: f the computing power does not improve, the effectiveness of the AI algorithm diminishes. I believe the chip makers are designing intelligent processors to optimize tasking, however the software and hardware deliverables need to be coordinated.
Q: What is the most exciting project you’ve personally worked on within the last 3 years?
A: I utilize Reinforcement Learning to validate the Agent tasking however I will design the Agent to build its own state diagram based on the patterns of data. This works very well in defining Zero day attack or image manipulation and various other problems because in these instances we are looking for the abnormal behaviour of the data “a needle in a hay stack”.
Q: What are you currently working on? What departments are involved?
A: I am driving AI in the following areas:
-COVID-19 mitigation software the detects face masks and social distancing
-Image optimization and corruption
Q: What can our audience expect from your session at the upcoming AI Accelerator Festival?
A: With the Artificial Intelligence, we can make our products smarter at defending against cyber attacks