Its billing model is consumption-based. I understand the consumption-based model, but it is not necessarily easy to estimate and guess how many points or how much we are going to consume on a specific application up until we get to that point. So, for us, it would be helpful to have more insights or predictability into what we can expect from a cost perspective if we are starting to use specific features. This can potentially also drive our consumption a bit more.

The other thing for us is that while it is great that we have all these standard metrics, it would be good if we can also more easily define standard metrics to be consumed for our own application. At the moment, for a lot of applications, we have to reinvent the wheel every time. If there was something so that we can build our own packaging of metrics, it would be helpful. In the future, we might be deploying our software to other customers as well. So, they should make it easier to redeploy that.

There should be more customizable dashboards. The Wavefront dashboards are very technical and a more business-oriented dashboard design would definitely help.

The documentation and integration with Kubernetes could be improved.

The main problem I have is that the license cost is very high.

Another thing is that, MSSP-wise, it would be better if we can add an MSSP as a separate product, to provide a security console integrated with Tanzu Observability, much like how the load balancer is integrated with Tanzu.

It's hard to set up Tanzu clusters. It's hard to do a POC. Once you set up a customer's environment, you easily see the problems.

The initial setup should be easier and more seamless. 

In the new version, I would love to see more prediction capabilities. It would be great if one could see the alerts get a little more enriched with information and become more human-friendly instead of the technical stuff that they put in there. I think those would be really awesome outcomes to get.

I would also like it if alerts were easier. It's really easy to drown in alerts. So what the developers of Wavefront could do, is to correlate different types of alerts to bring a much more enriched incident. That would be an interesting thing to look at. Predicting a problem beforehand could also be improved since Wavefront has all the data of metrics in its store. I would like to know if something is wrong so that I can go and look at it. I would like to use the AIML approach to predict problems for the user other than the users trying to figure out what the problems are themselves. I would love to see a change in that direction.

It could use a URL document server. Everything in the market is moving towards automation and everybody's looking for the single click operations as well relational data locality.

The virtualization could be improved.

I would like to see integration with Kubernetes cluster and APIs so that you can manage the entire stack.

Authentication is probably there already, but if it isn't, it should be added. Integration should be in there. They could make it more easy to plug-in data so that a nontechnical person will be able to use it, like accountants or finance people. That way they don't have to ask us.

The solution should integrate AI elements. Instead of opening a file with a template and filling things out, have the AI ask questions, and then let it fill it out. 

The implementation is a long process that should be improved.