ANOP: Autonomous Network Operations Platform

Autonomous Network Operations Platform (ANOP) enables Telecom Service Providers in evolving their Network & Service Operations to ANO Levels 3 & beyond, by delivering AIOps use cases, driving direct benefits of operations KPI improvement, through the following functional segments:  Cognitive Alarm Handler provides a normalized event arbitration framework that can drive outcomes using descriptive and predictive insights for alarm correlation, automated incident creation, and routing modules to reduce alarm backlogs and reduce MTTR.  Proactive Automation can help execute tasks to automate incident troubleshooting, diagnosis, and resolution tasks to assist remote troubleshooting actions and provide analytics-based field dispatch recommendations.  Smart Advisory functions provide prescriptive recommendations, including contextual guidance for incident resolution steps, scripts through adaptive operational dashboards or ingesting into tickets to resolve the incidents faster and on the first attempt.  AIML-based correlation of service level metrics with resource level metrics offers faster diagnosis and resolution of service level issues.  Latest Anomaly Detection models to monitor and identify anomalies in critical network & service KPIs/KQIs in near real time  TechM has built more than 200 use cases across the above functional areas and is developing another 100 additional use cases as a roadmap.

Key Characteristics of ANOP Framework and Guiding Principles Traditional operations practices shall evolve from process-driven to data-driven operations, by transforming tooling capability and operations environment. The following capabilities should be set to evolve into data-driven operations practices:

1. Data Foundation –Pipelines that ingest diverse data sets with governance and quality assurance across the Fault & Performance management lifecycle for multi-domain & heterogenous vendor deployed network
2. 360° Network & Service Fault lifecycle Observability –Organizing events/alarms/counters/KPIs/topology/TTs across multiple domains, and multiple layers of resources/applications and services
3. Performance Insights e-Focused Insights –Generating data-driven insights (via APIs) from a multiplicity of intelligence provider ecosystem, providing insights into network and services
4. Unified Automation Toolsets –Employing Classical AI/ML models with GenAI as well as Agentic AI based Use cases for AIOps, combined with workflow orchestrator module to orchestrate Incident, Change Management processes and performance function
5. Presentation Layer –Creating low code-no code dashboards and action hubs that can leverage insights for analytics, decision-making, and advisory purposes, customizable by Operations team and avoiding need for software coding skills

TechM’s deep domain expertise combined with AWS-managed AI infrastructure provides the flexibility to build comprehensive models, leveraging:

1. Amazon Bedrock to help create custom LLMs that are exclusive to the CSPs with the ability to use Agents that help transform insights to tangible actions and build a set of constantly evolving, live knowledge bases
2. Amazon Sage maker suite of AI/ML toolsets to build, adapt, and deploy AI per Life Cycle Management (LCM) requirements,
3. Amazon Managed Data Infrastructure that includes DynamoDB, Lake Formation, and EMR that grow with the demands and CSPs pay for what they use, and
4. Near real-time automation, helper functions such as Amazon Event Bridge, Step Functions, and Amazon Q Developer can assist in building actionable logic that follows CSP’s operational processes