I am currently working on the problem of modernization of legacy applications using techniques from program analysis and AI. The primary problem that I am solving is to recommend microservices out of the monolith legacy applications based on business functionality.
Prior to this, for around a decade, my primary research focus was on applying AI techniques in solving optimization and automation problems for IT operations. I have worked on a variety of research problems in the domain of AI for IT operations and service delivery including ticket classification, root cause analysis and resolution action recommendations. One of the notably high impact works was the automated data curation framework for bootstrapping conversational and QA systems. I used supervised and unsupervised machine learning and deep learning for this purpose. The conversation system was targeted for technical support to end users.