Architecting Enterprise Platforms for CI/CD, Cloud and SRE with AI and Analytics

Are machines ready to write, test, debug and patch enterprise software? What are the emerging capabilities? Is it a fantasy or fast happening already? How are enterprises strategizing and retooling?Gi …

Are machines ready to write, test, debug and patch enterprise software? What are the emerging capabilities? Is it a fantasy or fast happening already? How are enterprises strategizing and retooling?Given our rich experiences in consumer settings with intelligent voice assistants, recommender engines, facial recognition and sentiment analysis, we woud expect enterprise software platforms to be just as fortified with AI and ML capabilities where feasible. What can AI/ML do to enhance portfolio planning, CI/CD, Middleware/Platform/Cloud, SRE, Resiliency and CyberSecurity to name a few? How can we triage anomalies and reduce MTTR? There are indeed some fascinating developments. They are gaining traction amongst practitioners and they will be in action within enterprises sooner than we think. In this talk, we discuss the applicability of data science and analytics from portfolio planning, the development of safe software that has to be engineered to specifications, tested and deployed across a myriad of systems, devices and end-poinits both within and outside the firewall. The SRE team and customer engineering team have to provide an experience guaranteed to delight end-users. Specifically, we explore the parallels between agent-based learning systems where the machine mimics the experience of rewards/penalties in choosing the best path to reaching the end goal - which is to produce good quality software. We compare reinforced learning and hidden markov models with enterprise software engineering. This talk will be a survey of AI/ML tools enhancing enterprise software development and some of them will be demonstrated live.The authors will be sharing their code and future ideas as part of an open source development effort.