Skip to article frontmatterSkip to article content
Site not loading correctly?

This may be due to an incorrect BASE_URL configuration. See the MyST Documentation for reference.

Build production-ready anomaly detection systems using deep learning for observability data


This section covers advanced techniques for applying machine learning to observability data, with a focus on building production-grade anomaly detection systems using ResNet embeddings and self-supervised learning.

What’s Here

These tutorials guide you through building a complete anomaly detection system from scratch:

Topic

What You’ll Learn

ResNet Architecture

Adapting residual networks for tabular observability data — Understanding how to build custom TabularResNet models for time-series and structured data

Self-Supervised Learning

Training without labels using contrastive learning — Learn techniques to train embeddings on unlabeled observability data

Production Deployment

FastAPI services and vector databases — Deploy real-time inference systems with monitoring and automated retraining

Learning Approach

Production-First: Every example is designed to be deployable in production environments, not just toy demonstrations.

Hands-On Code: All tutorials are executable Jupyter notebooks with sample data—run them, modify them, build your own systems!

Complete MLOps: Learn the full lifecycle from data preparation through deployment, monitoring, and retraining.

Full Tutorial Series

For the complete, comprehensive guide to building anomaly detection systems with ResNet embeddings:

Observability Anomaly Detection →

The full tutorial series includes:

What You’ll Build

A complete production system including:

Why Observability + ML Matters

Understanding how to apply ML to observability data is critical for:

Prerequisites

These tutorials assume:

Target Audience

Where to Start

New to neural networks? Start with our Neural Networks From Scratch series to build foundations.

Ready to build? Dive into the full tutorial series and start building your anomaly detection system.


These tutorials use production-grade tools and techniques used by leading tech companies to monitor and secure their infrastructure at scale.