Observability Programmes

NSC-101

Observability Foundations

Instrument your first model endpoint with baseline metrics, structured logging, and health checks. Participants learn to define SLIs for inference latency and error rates, configure Prometheus exporters, and build a starter Grafana dashboard.

C$2,400 · 2 days · Up to 8 participants

• Metric taxonomy for ML endpoints
• Health check design patterns
• Starter dashboard template

NSC-201

Distributed Tracing for Inference

Deploy OpenTelemetry collectors across preprocessing, inference, and postprocessing stages. Learn span design for multi-model pipelines and correlate traces with structured logs for root-cause analysis.

C$3,200 · 2 days · Up to 8 participants

• OTel collector configuration
• Span context propagation
• Trace-to-log correlation

Prerequisite

NSC-101 or equivalent instrumentation experience.

NSC-301

Drift Detection Workshop

Configure automated drift detectors on input features, output distributions, and embedding spaces. Build alert thresholds with seasonality awareness and write incident runbooks for model degradation events.

C$4,800 · 3 days · Up to 6 participants

• Feature distribution monitoring
• Embedding drift analysis
• Degradation runbook templates

NSC-401

Alerting & Incident Response

Design intelligent alerting with deduplication, escalation paths, and on-call rotations. Integrate NeuralScope alerts with PagerDuty, Slack, and your existing incident management platform.

C$3,600 · 2 days · Up to 8 participants

• Alert fatigue reduction
• Escalation policy design
• Incident retrospective framework

Includes

30-day post-workshop alert tuning support.

NSC-501

Governance & Audit Logging

Implement immutable audit trails for model versions, deployment events, and human-in-the-loop decisions. Configure PIPEDA-aligned retention policies and generate compliance reports for risk committees.

C$5,200 · 3 days · Up to 6 participants

• Audit log schema design
• Access control integration
• Compliance reporting templates

NSC-601

Enterprise Observability Architecture

Capstone programme for platform leads rolling out observability across an entire ML estate. Covers multi-tenant collector design, cost governance, federation across regions, and organisational change management.

C$8,400 · 5 days · Up to 4 participants

• Multi-region collector topology
• Cost attribution frameworks
• Rollout playbook for 50+ endpoints

Deliverable

Enterprise observability architecture document and 90-day rollout plan.