Low 4G throughput
Daily Detection of cell-level 4G Low Throughput with automatic root cause analysis
Most of the Network engineers’ daily tasks (monitoring performance, troubleshooting top-offenders, optimizing clusters, fine tuning new rolled out sites) are highly manual, which usually means a lot of wasted time, inconsistent results and a high exposure to failure.
Time wasted in daily manual tasks
Inconsistency, inaccuracies and failure
Monitoring the network, detecting problems, and taking actions to fix them often require several iterations. Manually tracking all on-going actions, opening/re-opening/closing tickets are all low value-added engineering time activities.
Manually tracking all on-going actions
Low value-added engineering time
Networks dynamics such as capacity upgrades, vendor swaps, new customer traffic patterns, new technologies roll-out, legacy technology shutdowns, existing network densification, spectrum re-farmings, or operational incidents - make it difficult to keep an updated complete view of the network situation.
Difficult to keep an updated view of the entire network
Legacy technology and vendor lock-in
90%
More accurate
100%
Consistency level
90%
Manual effort reduction
2,5x
Staff productivity augmentation
Automatic actions can be directly triggered in closed loop for the most trusted and actionable problem categories, driving extra fast and zero-touch response to detected problems. “Ready-made” diagnostics and action plans organized and tracked via tickets which can be dealt with in open loop.
"Ready-made" diagnostics and action plans
Automatic actions triggered in close loop
Maintaining ML Models (MLOps & Govern) up to speed with data and network evolution. Digitalize their expertise (ML Models) and making it available to the whole organization. Driving automation by AI across the whole organization.
Digitalizing expertise (ML Models)
Driving automation by AI
Get started and request a demo to learn how Network Advisor can help you.
Daily Detection of cell-level 4G Low Throughput with automatic root cause analysis
VoLTE drop anomaly detection with automatic root cause analysis, discarding situations where degradation is transient
Accessibility anomaly detection with automatic RCA, discarding situations where degradation is transient
Congestion anomaly detection with automatic root cause analysis discarding situations where degradation is transient
Cell-level UL interference detection and classification model
Cell-level high proportion of Low SINR samples detection classified through a model
Inter-layer Imbalance detection with automatic classification root causes
Leakage 4G into 3G anomaly detection with automatic RCA, discarding situations where degradation is transient
RTWP anomaly detection with automatic RCA, discarding situations where degradation is transient
Cells' availability anomaly detection with automatic RCA, discarding situations where degradation is transient
It improves RAN PSF features' impact in your network. It constantly monitors the network correlating KPI degradations with PSF configuration changes, automatically proposing next best actions to mitigate impact while improving energy savings.
Below you will find answers to the most common questions about Network Advisor.
Network Advisor is designed to fit within an operator organization’s daily workflow on network performance and troubleshooting processes.
Network Advisor connects to relevant data sources, key ones being configuration, topology and network counters and KPIs.
The AI Engine correlates data from multiple sources to automate the detection, troubleshooting and action recommendations for network performance issues
Network Advisor SaaS is delivered in cloud service (e.g. AWS, Azure, etc.) and can also be deployed on-premises, in your private cloud or data center.
It is easy and quick to get started, fit for a faster procurement process, with a functional solution in operation within a week.
Monthly subscription. No strings attached. Stop at any time.
One key design principle of the system is flexibility, so as long as the proper data sources are integrated, and enough training samples are provided, any root cause that is relevant to you can be addressed.
It typically depends on a combination of targeted number of root causes, number of features required to model them and the number of available samples to train the system. Some of the models created by Network Advisor have reached over 90% accuracy.
Not at all. Network Advisor leverages unsupervised clustering techniques to detect group issues with similar patterns and helps you analyze as many problems as required to train your model in the most efficient way.
Network Advisor works in conjunction with other systems within your company, such as Ticketing, SON, CM, or Work Order management systems. Just as any engineer would do.
No coding skills or database modeling are required, only general concepts from data science are advisable to make the most of the system. Network Advisor relies upon TuplOS, Tupl's AI engine, which is a set of utilities designed to synthesize and digitize the complex knowledge base from your engineers and free them to focus on putting their expertise to work.
Tupl can guide you through the process via a coaching and training program from our services team.
Get started and request a demo to learn how Network Advisor can help you.
