Event Volume Anomaly Detection Solution Accelerator

tutorials/event-volume-anomaly-detection/introduction.md

@@ -0,0 +1,44 @@
+---
+position: 1
+title: Event Volume Anomaly Detection
+---
+
+Welcome to the **Event Volume Anomaly Detection** solution accelerator.
+
+This accelerator provides an introduction to detecting anomalies in Snowplow event volumes, specifically within **BigQuery** on **Google Cloud Platform (GCP)**. The focus is on identifying anomalies in **event volume data**, using data stored in BigQuery. This can help detect potential tracking issues or sudden increases in failed events by monitoring event volume trends.
+
+This guide walks you through the steps required to:
+- Load Snowplow event data into **BigQuery**
+- Train an **ARIMA+** model for time series anomaly detection
+- Identify statistically significant drops or spikes in event volumes
+- Visualize anomalies using **matplotlib** and **seaborn**
+
+![Application Output](images/anomaly-detection.png)
+
+## Requirements
+
+To use this accelerator, you need:
+- **Access to a GCP project** (including the project ID)
+- **BigQuery permissions** (read, write, and query access)
+- *A Snowplow pipeline is not required*, as sample data is provided.
+
+This accelerator typically takes around **30 minutes** to complete. The notebook requires minimal computational resources with the provided sample dataset.
+
+## Get started
+
+This accelerator is available on **GitHub**:
+- [GitHub repository](https://github.com/snowplow-industry-solutions/event-volume-anomaly-detection/blob/main/notebooks/bigquery/anomaly_detection.ipynb)
+
+You can clone this repository and run it locally on your PC or import it into a [Google Colab notebook](https://colab.research.google.com/).
+
+## Next steps
+
+Once you've completed this accelerator, you can:
+- Adapt the model to detect other anomalies in your own Snowplow event data
+- Fine-tune the confidence threshold to reduce false positives
+- Expand into detecting other types of anomalies such as missing properties or bot traffic
+
+By leveraging Snowplow’s granular event data, you can proactively monitor data quality.
+
+Ready to get started? [Jump into the notebook](https://github.com/snowplow-industry-solutions/event-volume-anomaly-detection/blob/main/notebooks/bigquery/anomaly_detection.ipynb) and start detecting anomalies!
+

tutorials/event-volume-anomaly-detection/meta.json

@@ -0,0 +1,5 @@
+{
+  "title": "Event Volume Anomaly Detection",
+  "label": "Solution accelerator",
+  "description": "How to build a real-time event volume anomaly detection system using Snowplow and BigQuery"
+}