It's been a while since we introduced the first version of Metrics, a powerful way of utilizing time series aggregations (or velocities) as part of your fraud and abuse prevention strategies. Today we're happy to release a fully customizable version of Metrics, which enables you to create time series queries that run in real-time, based on any fields available on events, including custom properties.

The top chart visualization can be a great way to analyze a slice of traffic, to get a breakdown of eg. the top countries or signals. To make it easier to see the relations between values we've added a new column which shows the percentage, as well as an interactive pie chart visualization which allows you to explore the data further.

Policies is a central feature in Castle for controlling automation and decisioning. In order to make it easier to see the impact of changing a specific policy, we've introduced a new view where you can easily compare how a policy triggers before and after your desired change. All this in the same powerful context of the Explore view. When you're happy with your modifications, you can quickly save the policy directly from this view.

Castle has supported webhooks for quite some time now, as a way of automating your fraud and abuse operations. We're happy to launch a much improved version of setting up webhooks and how they trigger. In the past Castle sent out webhooks based on two static policy conditions: one for when the policy action was challenge and one for when it was deny.

In the Explore view it's easy to quickly filter out interesting values that you may find during an investigation by simply clicking on them. We've now improved this experience by making the menu more compact and easy to read, plus added a new action for copying the value to clipboard. The latter was a highly requested feature, that facilitates cross-checking data between systems.

Today we're launching a new option to our Risk and Filter APIs: the ability to retrieve more details from the event context. By passing the expand: ["all"] option as part of the JSON payload you're sending to either the Risk or Filter API, you'll get back a rich object with additional data such as IP-location and device information. This information is the same as you can see in the Castle Dashboard when exploring data, but now you can get it directly in the API response as well.

We're happy to announce that we have introduced several dashboard improvements for Lists.

Today we've released a new option to filtering timestamps - the ability to compare a timestamp to the current time, in order to e.g. find users that signed up "in the last 2 hours". This filter opens up the possibility of creating Policies that flag users that signed up and then quickly did something suspicious, like creating 10 new content posts. Seeing high velocity abuse from new accounts like this is very common across multiple types of services.

We're excited to release our updated onboarding flow for new Castle accounts. There are many ways to start using Castle, so to make it more streamlined we've listed the most important steps to create visibility into your application and get a good foundation for protection in place. This includes default Lists and Policies for common cases of fraud and abuse, as well as recommendations for notifications.

We've made it easier to create a filter that targets one or more specific Policies. This filter is leveraging the same tab-system that we introduced for Lists and Signals a while back. With this change it's easy to evaluate a specific Policy based on historic data and get an overview of how many users or devices (or anything else) were affected.