Most networks, in both enterprise and telecom, are constantly evolving, and with the advent of 5G, the enterprise-telco boundaries are coalescing. Monitoring the network KPIs does not reflect the consumers’ experience. Network observability, beyond just visibility, is the key driving factor in providing the best user experience. Network operators must know if the consumers, such as mobile devices, IoT endpoints, or subscribers, experience continuous service usage. They will also have to ensure subscriber affinity and retention. To provide an optimum network performance, the provider should be able to assimilate logs, data, network packets, etc., from disparate sources and provide a deep insight (AI/ML) derived from these data collection points. PrObserv is Telaverge’s programmable open source-based cloud-native network observability framework. From a single visibility platform, it enables capturing, analyzing, troubleshooting, and reporting across a wide range of protocols. It provides dashboards to visualize, debug and arrive at actionable insights. It uses Elasticsearch, Logstash, and Kibana (ELK) stacks and a programmable API for data normalization and packet parsing.

Customer pain points

Maintaining high-performance and complex heterogeneous networks for anomalies could be challenging for most Service or Network providers. The task gets even more daunting for telecom networks with call flows spanning multiple domains and protocols with data flow limitations, complicating troubleshooting issues, and adhering to Service Level Agreements (SLAs). It encompasses integrations from different sources, capturing, storing, filtering, parsing, and troubleshooting the applications running on distributed infrastructure. This requires advanced telecom protocol expertise, and the capability of stitching call flows across different nodes and packets. Telaverge’s PrObserv streamlines this complex process and facilitates a simplified solution to address this problem and other related pain points.

Telaverge's PrObserv Features

PrObserv can ingest, collect, and integrate data from sources like log files, Call Data Records (CDRs), PCAPs, raw packets, Kafka streams. etc. The information is then processed and stored at the backend for further analysis.
Telaverge’s PrObserv is extensively scalable. It’s built on top of an ELK stack with the Data Processing Engine (DPE), which can process terabytes of data in near real-time. The DPE is a programmable entity, providing the ability to customize and add new data parsing capabilities through APIs.

The dashboard provides insights and presents a consolidated view of statistical data, network health, traffic, patterns, and historical data.


The dashboard is customizable with pre-configured templates for various graphs and tabular views to aid users in visualizing data and drilling down to details for specific information. These dashboards can be modified to consume analytical data produced by the AI/ML modules.


The troubleshooting section provides detailed diagnostic analysis by filtering information from data based on protocol-specific parameters and displaying the visual call graphs that provide actionable metrics. Different filtering mechanisms allow the call graph section to show the flow of data exchange across network elements.


Additionally, PrObserv can deploy Machine Learning (ML) models to identify anomalies and patterns in user and element behaviors. For example, clustering algorithms can be applied to find if a significant shift in subscribers’ data usage is correlated to an event, like changing customer handset, data plans, network outage, etc. It can also isolate anomalies in call flows, peak, average traffic, etc.

PrObserv has industry-standard notification and alerting mechanisms. This enables users to be notified on preferred channels like email and SMS on event triggers with thresholds or to receive periodic reports. These events and the corresponding actions are configurable, empowering users to customize notifications and alerts.


PrObserv is built on open-source software like ELK to analyze unlimited data. The DPE uses standard libraries and packet capturing tools like libpcap for Wireshark and Tshark, to listen and analyze data packets. It uses ML libraries like Numpy, Pandas, Matplotlib, etc., to perform the logical functions. This platform can accomplish Extract-Transform-Load (ETL) operations, identify anomalies, and many other features. PrObserv is a cloud-native software. It provides the advantage to process terabytes of data, reduce latency, and offer scalability to support massive network traffic. It is easy to deploy and provides various other PaaS-based advantages over traditional models. Additionally, per customer requirements, it is deployable on-premises using Bare Metals (BMs), Virtual Machines (VMs), and Private or Public Cloud  setups.  


Supports a wide range of protocols such as TCAP (CAMEL, LIDB), DIAMETER, RADIUS, SIP, etc. The architecture is extensible to allow adding new protocols with ease, making it a protocol-agnostic solution.


Telaverge’s PrObserv offers numerous advantages like

  • PaaS-based solution.

  • One platform – capture, monitor, analyze, store, troubleshoot, alerts/reporting, and custom business logic.

  • Protocol agnostic, high performing and massively scalable elasticsearch based.

  • Flexible dashboard and extensible post-processing with ML-based pattern and anomaly identification.

  • Massive cost advantage over similar offerings.


Telaverge’s PrObserv addresses most customer challenges in managing their networks without investing in multiple products, vendors, or dedicated teams.

PrObserv is a future-proof ML- based enriched carrier-grade monitoring software that gains actionable insights for Any Vendor, Any Network, and Any Service.

Telaverge, a global communications and technology provider, helps customize and simplify solutions for you. For more information on our products and servicesemail us at