AIOps monitors and assures services to reduce risk, improve customer experience, productivity and business continuity
Robi Axiata wants to manage and resolve customers’ complaints from end to end, with little or no human intervention. It is also striving to improve customer experience by minimizing complaints about and the duration of issues that impact service level agreements (SLAs). It does this by automating repetitive operational tasks and incubating awareness of operational activities through an AI-powered backend. The solution uses machine learning and deep learning as part of natural language processing (NLP) and understanding, predictive analytics and decision making, using both supervised and unsupervised models.
Robi Axiata’s AIOps solution contains three key components, complaint, performance and capacity management. It provides a human-machine mid-layer interface that analyzes data from communication channels, structured or unstructured log files, calls detail records and bot-generated reports to capture information from warnings, alarms, health metrics, security issues and customers’ complaints. The solution also draws on system and service metrics, account management and event triggers. Then it generates remedial actions, on the fly, derived from robotic process automation (RPA), and provides feedback on the success of the remedy to refine future responses.
e.g. fewer incremental human resources; annual maintenance costs cut as most Level 1 & 2 problems fixed by AIOPS
will increase with addition of new launch platform
from digital CRM to the application provider
from overall monitoring
Digital Service, IoT & VAS, IT
“Most IT and digital platform operations time is spent troubleshooting complex system and customer problems across heterogeneous solutions. This isn’t efficient in a challenging business context. AIOps play a pivotal role in collecting all structured and unstructured IT and network data, analyzing event analysis, KPI-based performance data, log pattern searches, and more. AIOps can detect anomalies, recognize and forecast incidents, and proactively solve customer complaints to provide channel-agnostic fulfillment.”
Deep learning and neural networks
The backend is powered by machine learning and an offshoot of it, deep learning. Machine learning uses processes like predictive models, while deep learning uses artificial neural networks which are designed to imitate how humans think and learn.
The AIOps solution interprets, triggers responses, and communicates with other systems to perform a vast variety of repetitive tasks, resolve issues and give feedback, equipped with a fully automated continuous integration/continuous delivery (CI/CD) toolchain powered by a modular microservices-oriented architecture.
The operator summarizes the benefits it gained as better risk mitigation, customer experience, decisions and productivity, plus failsafe business continuity.
By signing up for this case study, you agree to the Privacy Policy