The Loft

Chair: Sanjay Mewada, Analyst & Chief Research Officer, Recon Analytics

12:00 - 12:30 
The Chain of Trust: From Governed Data to Autonomous Action
72% of CSPs say their AI is trustworthy. Only 14% can prove it. This session introduces a six-link Chain of Trust:
Source → Governed Record → Reconciled State → Causal Inference → Authorised Action → Verified Outcome

Part I: Creating Trusted Data
Source → Governed Record
Trust begins with evidence-bearing data. Learn how provenance, encryption, policy enforcement and tamper-evident audit trails create a governed record that can be trusted and verified.
Featuring: AXON Datum in a live sovereign-government data platform.

Part II: Delivering Trusted Autonomy
Reconciled State → Authorised Action
See how trusted data is transformed into trusted decisions. AXON Datum, Cortex, Digital Twin and Neura combine to reconcile network state, infer causality and execute TM Forum's observe-predict-act-verify loop (AN-4).
Using a live CSP workflow, this session demonstrates why autonomous networks require not just reconciled data, but provable data.

Siddhartha Dattagupta, CTO, AXON Networks
Kumar Vishwanathan, CPO, AXON Networks 

12:30 - 12:45
On-demand NOC – Extra support when operations teams are overloaded

NOCs are built for managing common incidents and outages.  But major network outages — when incidents spike, alarms surge, or the team is short-staffed — can lead to significant problems with customers. Adding permanent staff or infrastructure is not realistic. At these times, CSPs need solutions for adding trusted operational capacity, fast.

This 15-minute use case discussion explains how NOC as a Service serves as a practical on-demand model for those overload moments. The session will discuss how a cloud-based agentic AIOps and automation solution can be quickly integrated and activated when needed, giving network operations teams more capacity during critical outages and scaled back when the network stabilizes. 

• Where NOC as a Service helps most, and how to use it during incident spikes, alarm surges, major events, or staffing gaps.
• What to measure first, including faster triage, fewer escalations, reduced downtime, lower truck-roll risks, and avoid overbuild.
• How public cloud-based solutions enable the on-demand NOC model, making it easier for CSPs to leverage analytics, automation, and elastic compute when pressure rises, then scale back after the event. 
• Start with a straightforward NOC use case – for example, incident detection or root cause analysis – evaluate the outcome and then expand onto other use cases. 

Hitesh Agrawal, Global Alliance Lead, Telco, Amazon Web Services (AWS)
Ram Ramanathan, VP Product – AI, Automation & Cloud, Ribbon Communications

12:45 - 13:05 
Project foundation - AI-native ODA canvas reference 

• Introducing the Foundation AI Native ODA Canvas reference demoed at DTW 2026
• Exploring how cloud, data, and AI services combine to support AI-native business architectures
• Demonstrating reusable patterns for building intelligent enterprise solutions
• Highlighting collaboration between platform, data, and AI ecosystem partners

Jamin Solensky, Lead Solutions Architect, Databricks