
Anne-Flore Agard
Head of Telecom at Capgemini Engineering

12:45 - 13:15
Grounding the agent: Why agentic network operations start with data, not models.
Most AI-driven network automation deployments stall before achieving a fully closed loop not because the models are inadequate, but because the data layer they operate on was never designed for autonomous action. This session makes the case that agentic network operations is fundamentally a data architecture problem. Drawing on deployment patterns across Tier-1 operator environments, we identify four precise failure modes that block AI closed-loop execution: unresolvable state across conflicting OSS systems, broken causal chains between alarms and remediation paths, absent post-change verification loops, and temporal data skew — the often-overlooked problem of AI agents reasoning over composite snapshots assembled from systems with wildly different polling cycles and propagation delays.
A practical four-stage insertion roadmap shows how operators can deliver measurable value within 90 days while building toward full domain autonomy and how reaching AN-4 unlocks the shift from connectivity provider to Operator-as-a-Service.

Head of Telecom at Capgemini Engineering


Senior Vice President of Global Sales

Chief Technology Officer



Discover how operators build governed, explainable networks that think, protect, and evolve independently. Explore semantic digital twins as ODA-aligned foundations for closed-loop automation and AI-driven decisions at production scale. See MODaaS demos proving model trustworthiness, AI-native security agents enforcing policy-driven compliance in autonomous workflows, and unified data platforms stopping fraud while powering breakthrough experiences.




What changes in accountability, assurance, and leadership when autonomy becomes real.

Learn how teams overcome real-world integration bottlenecks.


This DTW Ignite Lunch Briefing marks the launch of the DT4DI White Paper 4.0: Leveraging AI-Native Architectures to Empower Trustworthy AI. Designed for a select group of senior CSP executives and industry analysts, the session will explore how to move from experimentation to scalable, trustworthy AI deployment.