Building Robust Data Pipeline with Apache Spark
Resilient Spark pipelines, data platform reliability, and operational design.
Speakers: Zak Hassan
Speaking Archive
Technical talks from the intersection of SRE, data platforms, Kubernetes, and AI operations.
A curated wall of past presentations covering Apache Spark reliability, Prometheus monitoring, OpenShift-native ML workflows, GPU observability, MLFlow, and anomaly detection.
talks.indexed7
primary.topics7
archive.range2017-2019
Presentation Signal Wall
Past sessions, organized like an ops dashboard.
Each talk is a different telemetry stream: distributed data systems, Prometheus, ML platform operations, GPU monitoring, and anomaly detection.
Scalable Monitoring Using Prometheus with Apache Spark
Metrics architecture for Spark workloads and high-cardinality operational signals.
Speakers: Zak Hassan and Diane Feddema
MLFlow Model Lifecycle Manager and Kubernetes Operator
Experiment tracking, Kubernetes operators, and reproducible ML workflows.
Speakers: Zak Hassan and Mani Parkhe
Log Anomaly Detector Service
Anomaly detection patterns for noisy logs and operational triage.
Speakers: Zak Hassan
Monitoring of GPU Usage with TensorFlow Models Using Prometheus
GPU telemetry, TensorFlow workloads, and Prometheus-based visibility.
Speakers: Zak Hassan and Diane Feddema
Unsupervised NLP for Log Anomaly Detection
Using unsupervised language techniques to surface suspicious log patterns.
Speakers: Zak Hassan and Michael Clifford
Introduction to Hyperparameter Tuning in Machine Learning with MLFlow
MLFlow operations on Kubernetes-native infrastructure.
Speakers: Zak Hassan and Hema Veeradhi
Compare Notes
Want to discuss SRE, observability, Kubernetes, or AI infrastructure?
These talks are older public artifacts, but the operating themes still map directly to how I think about platform reliability, telemetry, and infrastructure leverage today.
Connect on LinkedIn