)
Data - value
cognitive networks
Telco and Enterprise companies rely on easy access to data. Valuable insights can be automatically extracted from the network itself with analytics tools and Artificial Intelligence. Reply can help Network Operators to build their business strategies and decisions based on real value network data.
the telcos' challenge
How to collect and analyse large amounts of data, identifying the useful insights and creating an informed and effective decision-making process?
Skills and technologies needed
To leverage collected data, companies need skills and technologies that guarantee real results in terms of operational and business benefits. Moreover, Machine learning and AI use cases have shown great potential in assisting with anomaly detection, root cause analysis, managed services, and network optimization. New technologies such as deep learning and reinforcement learning can be used to automate the network design process and optimize network performance in real time.
But to work effectively, all these applications require specific computational, pipeline and support infrastructure. The need can be addressed with data infrastructures that ensures scalability, speed, security and flexibility, acting as the foundation of the entire Data ecosystem. The goal is to trigger a virtuous process: derive more value from data and identify useful insights to activate action on the networks in a timely and proactive manner.
The solution
On this data foundation, you can implement effective and efficient data management for different types of data, qualitative and quantitative, on which to build custom models capable of adapting to the changing needs of the business. The data will be ingested from distributed databases or streaming data pipelines. Across the end-to-end pipeline, data will be versioned, tracked, and validated.
Similarly, multiple versions of ML models will be implemented and managed by a model registry and a model management workflow. To deploy models in various environments, workflow management tools will be used to manage the entire end-to-end flow. These pipelines can be implemented using cloud native ML workloads, deployed as container-based micro services, and their infrastructure footprint ranging from a single node running Docker to a cluster of nodes managed by a container orchestration system, such as Kubernetes.
Network Data Quality Processes
Complex organizations like Telco Companies co-exist with systems and applications devoted to specific functions, where databases are "isolated" from one another, redundancy is high, and data is often misaligned. The Network Data Governance is a set of strategies, processes, and rules for managing and enhancing data to increase the corporate data value of data-driven businesses. The objective of a Data Governance system is to use and improve processes for preventing and correcting issues with data to improve data quality available for decision-making. There is a need for a simplified view of the whole. A unified and certified view of the data, compared to the various sources that make up the information assets of a company.
A network analytics data governance system could be a useful tool to support long-term business competitiveness for Telco Operators. It provides:
- Centralized monitoring and general governance of data flows
- Greater efficiency for data control
- Unique interface for system owner, user line, and application management groups
- Real-time standard controls for each type of data stream
- Advanced customizable controls
- Easy update process (new streams, time shift, etc.
- Immediate evidence on data flow issues
Use cases of interest
Following our customers interests and market directions, Reply explored different areas of applications for Advanced Analytics, ML and AI techniques.
Fault management and prediction
Service disruptions and faults are inevitable in a telco network, so this is a critical area in which AI can play a key role. Applications in different domains and focus on different devices will be considered accordingly to our customers’ needs.
Traffic prediction and optimization
Network traffic prediction and optimization are required from customers aiming for CAPEX efficiency improvements and QoS enhancement. The telecom sector is driving towards self-optimizing networks based on real-time and event-based data, which automatically react based on inputs provided with ML/AI algorithms.
Churn prediction and customer care
Telco companies pay close attention to their own end customers. The information gathered from both the OSS and BSS worlds can be properly manipulated to prevent breakdowns, QoS losses and complaints from the end customers of a Telco provider. The Assurance and the Customer Care departments are focused on trouble ticketing opening prediction and churn prevention, which are now widely explored thanks to AI techniques.
Advanced analytics and visualization
The development of Advanced Analytics solutions, combining Data Mining and Pattern Recognition techniques to the domain knowledge, are vital for customers with interests in leveraging the hidden potential of the data and with interests in supporting the network troubleshooting. Root Cause Analyses can be pursued with these solutions and can lay the foundation for Fault Management Use Cases. Deeper investigation of the visualization tool will support our customers in making data-driven decisions and in exploiting advanced reporting solutions.
Image/Video elaboration and object detection
With the development of 5G networks and the growing use of drones, the ability to manipulate images and videos properly to gain useful information, which can processed at the edge, is becoming a fundamental need for our customers.
Why Reply?
Reply is committed to a continuous process of researching, selecting, and promoting innovative solutions on the market, capable of supporting the creation of value within organizations.
In recent years, the awareness of the vast possibilities deriving from the exploitation of Network Data has increased. For our Telco customers, we aim to combine technological skills of data analysis, modelling and process reengineering in order to favor the activation of a real and concrete path of cultural change and introducing a new approach to the issue of data management, whether internal to the company or deriving from external agents.