Elias Schwalme

Led cross-functional squads as acting Product Lead, leveraging Python-first microservices, SOLID/TDD practices, and GitOpsdriven Kubernetes (Helm & Argo CD) deployments; applied proven telecom transformation playbooks to deliver every milestone ahead of schedule and under budget. Established real-time KPI dashboards and model-drift monitoring on Databricks Model Serving with Prometheus/Grafana observability, providing C-Suite continuous 99.99%-uptime visibility into ROI, risk, and compliance status.

Implemented a Kubernetes-based monitoring stack Prometheus, Grafana, OpenTelemetry, and Evidently AI on GPU nodes fronted by Triton Inference Server, with Argo Rollouts and KEDA for auto-scaling. This cut our mean time to detect incidents. Worked closely with the trading and grid-operations teams to pipe drift and KPI alerts into Slack and Splunk, tuning thresholds to match live P&L and grid-stability metrics.

Set up SHAP and LIME dashboards in Grafana, added Evidently AI panels for drift monitoring, and fed everything through Prometheus and OpenTelemetry. Immutable logs now stream to Splunk and archive in S3 Glacier, giving auditors round-the-clock, SOC-2-ready traceability. Built an AI compliance framework covering the draft EU AI Act, TISAX, and ISO 21434, introduced model-risk tiers, and automated audit logging and explainability reporting.