Jepto — Anomaly Detection with Machine Learning

Never worry about missing changes in data

With lots of data, comes lots of alerts and no action. We give you the power to control what metrics are being tracked so when something happens you can take action.

You're in Control

Select the metrics that will be monitored for anomalies and avoid the noise.

Adjust the Sensitivity

Avoid getting alert fatigue by increasing or decreasing the sensitivity.

Positive and/or Negative

Choose if you want to be notified about positive changes, negative changes, or both.

Take Action

Choose who and how you want to be alerted, so you can take action before any damage is done.

External Workflows

Automate task creation in your Project Management tool or internal system through our Zapier connector.

Apply in Bulk

Drag and drop functionality to apply a monitoring template across multiple Clients. (Coming Soon)

Sophisticated Algorithm with Easy Setup

Anomaly Detection requires a lot of complicated math to ensure that you don't get alert fatigue from false positives. We make the whole process simple, just select the metric you want to monitor and set your notification preferences. It's that easy.

Want to tweak the anomaly detection algorithm? We got you covered with live previews of the last few months, so you can see how many anomalies you would have been alerted to. Adjust the algorithm settings and the look-back preview automatically updates.



1. Select a Metric to Monitor



2. Setup Notifications



3. Start Monitoring



4. Acknowledge



5. Resolve

Anomaly Insight Card

When an anomaly is detected an Insight card will be created that gives you all the information you need to understand the anomaly and what actions you can take next.

All items in Jepto have a similar card layout that gives you the most critical information in an understandable way. Here is the breakdown of an anomaly insight card:

The data source of the anomaly and the Client name (if applicable)
Title of the card
Summary of the card, with the metric and Segment (if applicable)
Visual indicator to quickly understand if the anomaly is a decrease or increase
Date the anomaly occurred
Overflow menu contains shortcuts and the ability to edit the Anomaly configuration which created the insight
The expected value based upon your previous data
Actual value that is considered to be anomalous
An easy-to-understand description of the anomaly data
When multiple anomalies are created over time you will be able to see them in a list, so you can switch between them or decide the
Button to open a draw to give you a detailed view of the anomaly
Acknowledge the anomaly to let others know that you are aware of it
Resolve the Insight to dismiss the card and inform others that the anomaly has been investigated and resolved
Comment on the Anomaly to communicate with other teams and keep a record of what caused the anomaly.