Solution Brief
AI and Machine Learning for Network Analytics
Outrank competitors and optimize ROI with fast, AI-driven automation for superior network quality and services
The issue
More communications service providers (CSPs) are now using artificial intelligence solutions to reach tier 4 (highly autonomous networks) of TM Forum’s Autonomous Networks. By automating end-to-end service operations and improving network planning with Smart CapEx, AI solutions help CSPs enhance customer experience throughout the entire customer journey. In addition, AI delivers better network quality and scalability, improves campaigns, reduces churn and, most importantly, brings in new revenue. That’s especially true with 5G premium services for businesses and consumers. New revenue requires AI technology solutions – including machine learning (ML), natural language processing (NLP), generative AI (GenAI) and large language models (LLMs) – and the automation they make possible in managing a 5G network’s complexity.
However, GenAI and LLMs by themselves don’t create value. That requires taking the customer viewpoint to make network and service improvements. To do that, CSPs need to integrate network and customer data, optimize that data and use full DataOps, ModelOps and DecisionOps to get the most from data analytics and AI. The right telecom data analytics and AI tools result in increased automation along with the opportunity to reach TM Forum’s tier 5, or fully autonomous networks. Greater automation delivers trustworthy insights faster, manages complexity and improves productivity, leading to better business outcomes and the agility required in today’s market.
The challenge
Improving network quality and service
Being competitive in the telecom industry requires better understanding subscribers, meeting their demands, anticipating their needs and making significant improvements proactively.
Complexity
Continual network changes, exponential data growth and customer behavior changes all lead to challenges with network anomalies, planning issues and increasing costs.
Budget constraints
Without the money to invest in network digitalization and automation, customer satisfaction and market share decrease.
Innovation
Developing new services and providing better customer experiences in an ethical, trustworthy way while becoming the best network is difficult without the right tools and insights.
SAS AI for telecom helps CSPs consistently achieve ROI improvements, including:
(up to 15% revenue growth.; 25% more accurate network planning forecasts.; up to 95% automation of operational forecasting processes.; 10% more satisfied customers.; up to 30% fewer customer service calls.; 12% less service-related customer churn.; up to 3-5% more asset availability.; 25% fewer unnecessary truck rollouts.; up to 5% savings in overall energy costs and carbon dioxide output.)
Our approach
SAS offers a scalable, open platform to manage diverse data across all network systems, integrating different approaches, coding languages and techniques to accelerate business transformation. Enjoy superior performance and resilience with secure governance and centralized administration.
We approach the problem by providing software and services to:
Deliver exceptional network quality and service
Maximize network quality and drive end-to-end service assurance in the current network, plus 5G and beyond. Trustworthy AI provides real-time insights for network management and operational improvements.
Reduce costs across the business and empower users
Save millions by quickly detecting, predicting and resolving issues. Transition from siloed data and operations to enterprise collaboration on cost savings and more.
Maximize ROI and improve customer satisfaction
Increase ROI with investments in network performance improvements that will boost your bottom line and excite your customers. CSPs that go with SAS AI for telecom improve their net promoter score (NPS) by up to 30 points and reduce churn by up to 50%.
Innovate and scale faster, maintaining higher profitability
Speed launches of new services and revenue streams that will keep your business profitable and improve productivity while delivering the products and services that help maintain customer loyalty.
Attain compliance with upcoming AI regulation acts
Implement the embedded trustworthy and ethical AI approach across data, AI and decisioning life cycles.
SAS difference
Outrank the competition
Fast, AI-driven automation delivers superior network quality reliability and services while optimizing ROI.
With SAS telecom data analytics and AI technology solutions, you can:
Integrate seamlessly
- Work in your preferred coding language or use SAS’ low-code/no-code environment to score models anywhere (database, memory, stream).
- Manage customer and network data at scale, add real-time analytics and AI capabilities for accurate recommendations on operational efficiencies and improvements, and ensure superior performance with secure authentication, governance, regulatory traceability and centralized administration.
Optimize network investments and uncover avoidable spending
- Use AIOps and MLOps for transparent insights to enhance network planning, deployment, maintenance and optimization.
- For greater service assurance, align latency and bandwidth with critical workloads using traffic forecasts and what-if scenarios.
- Advanced multihierarchical forecasting helps prevent unnecessary spending by identifying where capacity is needed and where it isn’t.
Manage large-scale network automation and create a self-healing network
- Create end-to-end network life cycle automation (NLA) for autonomous, zero-touch operations across multiple domains – including the RAN, IP transport network and mobile core network – and vendor solutions.
- Use intelligent automation to detect, prevent and resolve anomalies before they occur, minimizing costs and ensuring network reliability.
Unlock subscriber insights to generate greater customer value
- Gain real-time, actionable insights into customer intentions, needs and events that could affect experience.
- Apply NLP to LLMs to retrieve useful feedback from call center transcripts, surveys, social media and more.
- Combine network and customer insights to speed personalized product and service design.
Invest for maximum revenue and create new ROI
- Innovate faster with AI, machine learning and GenAI, creating additional revenue by investing in revenue opportunities such as network slicing.
- Increase ROI by supporting and accelerating new service launches and revenue streams, as well as intelligent forecast planning that prioritizes network rollout.
Use cases
Network automation of a network operations center (NOC)
A tier 1 CSP in Italy used SAS for its transformation from a rules-based NOC to an AI-driven service operations center (SOC). By managing its networks with customers in mind, it prioritized alarms that directly impacted customers and deprioritized those that impacted its network but did not affect customer service and predicted failures before they happened.
Model creation and deployment for improved data preparation and new revenue
A tier 1 CSP in Norway deployed SAS technology to create a models factory consisting of ModelOps, DataOps and DecisionOps. That cut data preparation time by 93.75% for new data science projects and created significant new revenue.
Value-based network planning (Smart CapEx)
- SAS advanced telecom data analytics and AI prioritized RAN and fixed-line fiber (FTTX) network rollouts or capacity upgrades based on ROI. Using AI and machine learning, a group of tier 1 CSPs in Spain and a mobile CSP in Africa saved an estimated 3-5% of their overall network RAN capital expenditures budgets. To create the savings, they generated highly accurate (95%) traffic forecasts for microlocations over 12-18 months, a level of forecast accuracy required for upcoming 5G deployments that have low radio coverage in urban areas.
- Another fixed-line operator in central Europe estimated from a pilot they conducted that it would reduce its fiber rollout planning from days to minutes by using SAS AI and ML to reduce much of the manual work.
Energy optimization
Using the traffic forecasting principle and intelligent RAN cell/site shutdown, a tier 1 operation in the Nordic countries ran a pilot that saved 5-25% on energy consumption from its 4G RAN network. Savings varied by time of day, year and number of cells/sites recommended to switch off.
Recurring trouble-ticket prediction
A converged tier 1 CSP in the Middle East had challenges with its copper and fiber fixed-line business, so it used AI-powered telecom analytics from SAS to predict issues that were likely to be reported repeatedly and prioritized treatment for VIP customers by assigning them the most experienced customer service agents.
Real-time network anomaly prediction
A tier 2 CSP in South America used streaming data to predict how weather and power frequency and voltage oscillation affected IPTV and broadband service quality in terms of packet loss and jitter. It sent proactive messages to its customers before they experienced service degradation, improving transparency and customer satisfaction.
Reduced bill shock
Research from the TM Forum catalyst project showed that introducing AI in telecom for predictive bill shock and GenAI for bill analytics reduced customer service billing inquiries by 15-26% and increased revenue 3-5% by improving customer satisfaction and revealing upsell and cross-sell potential.
Telecom call center optimization and next best offer/next best action (NBO/NBA)
Flexibly manage your telecom call center’s schedule to forecast traffic, optimize and manage correspondence and complaints processes, automate the analysis of conversation transcripts, and use insights to improve communication and prevent customer churn. Define your own integrations with LLMs for niche use cases. The value of adding SAS GenAI includes:
- 8–15% cost savings in handling complaints.
- Up to 20% more complaints handled.
- 20–40% faster average complaint handling time.
- 30–40% faster average response time.
- 20–25% faster complaint resolution time (complaints also decreased).
- More consistent customer experience with increased personalization.
More efficient new telecom product and service development
Keep telecom customers engaged by delivering the personalized products and services they want and need. GenAI helps create virtual designs and simulations, early research analysis, improved quality and testing, and much more.
-
Deliver exceptional network quality and service
Maximize telecom network quality and drive end-to-end service assurance in the current network for 5G and beyond. Trustworthy AI in telecom provides real-time insights for making operational improvements.
-
Reduce costs across the business and empower users
Analyze network performance and save millions by quickly detecting, predicting and resolving issues. Prevent unnecessary spending by showing planners where capacity is needed – and where it isn’t – with advanced, multihierarchical forecasting.
-
Maximize ROI and increase customer satisfaction
Telecom network analytics help you improve investments and performance. SAS helps TSPs improve their NPS score by as much as 30 points and reduce churn by up to 50%.