Sahil Yadav is senior director of product management and head of software and AI at Applied Optoelectronics, where he develops self-healing broadband networks. His work focuses on making automated infrastructure reliable enough to detect faults, recommend repairs, and take corrective action without sacrificing human oversight.
His career has included IBM, network automation at Cisco, predictive analytics at GE, and connected-worker safety at Guardhat. He later became chief product officer at TelemeTrak before joining Applied Optoelectronics.
At Guardhat, Yadav helped build systems that combined wearable devices, environmental and health telemetry, and real-time analysis to protect industrial workers. Inaccurate GPS data produced false safety alerts, and workers began ignoring warnings. That experience made GPS drift and alert fatigue central to his thinking: dependable automation requires validated inputs, traceable decisions, and intervention when confidence breaks down. He explored those problems alongside Hariharan Ganesan at AI Engineer World’s Fair 2025.
- Policy-constrained network autonomy: Automated broadband repairs should begin with reversible, low-risk actions, then expand under explicit operational policies, human approval thresholds, and auditable decision trails.
- Telemetry-to-remediation feedback loops: Autonomous network optimization links continuous monitoring, anomaly detection, diagnosis, and corrective action while escalating sensitive decisions to operators.
- Automation that respects legacy infrastructure: Self-healing cable networks must accommodate older equipment, unreliable telemetry, regulatory constraints, privacy requirements, and fair service prioritization.
At AOI, these priorities inform QuantumLink network-management software, which combines predictive diagnostics, infrastructure telemetry, and operational workflows for broadband providers.
