Satya Pradan, Cisco Systems, USA
Abhishek Surana, Cisco Systems, India
Venky Nanniyur, Cisco Systems, USA
Today’s cloud environment requires services to be deployed directly to the cloud and expect only hours of turnaround time for critical software fixes. Customers have been demanding for shorter deployment cycle even for traditional on-prem deployments. This has created challenges as well as opportunities for software delivery and deployment teams to predict quality during development activities and take corrective actions before the software is deployed to the cloud or released to customers. At the heart of this challenge is the prediction of software quality before the software is released to customers. This poster presents a quality prediction model based on machine learning for large enterprise-grade software. Application of tphis work to Cisco’s IOS-XE based software releases shows tremendous opportunity for using ML/AI based modeling to predict and improve the quality of large-scale enterprise grade software systems. Our motivation is to share the details of a real-life application of ML/AI model to enhance software quality and encourage other companies to share similar results for large scale software.
Learning:
- Challenges with quality prediction modeling.
- Predicting software quality using AI/ML.
- Using dominant attributes to improve software quality
- Increasing the success of software deployment