Intelligent observability: Kloudfuse introduces AI-enabled platform upgrade

Bart: So, who are you? What's your role, and where do you work?

Michael: Hi, I'm Michael Wilde. I'm the head of sales at Kloudfuse.

Bart: What do you want to share today?

Michael: Glad you guys stopped by the booth. Kloudfuse is a really awesome observability platform. We've been in the business for a couple of years, and last week, we came out with the third release of our product, which includes a whole bunch of new innovations. Folks that use Kubernetes will really love Kloudfuse for a couple of reasons, and I'd love to tell you about that in just a minute.

Bart: What problem does Kloudfuse solve?

Michael: Kloudfuse is a full-stack observability solution. When you think about all of observability, there are a ton of data sources. We've got logs, metrics, and traces. What about the stuff from Kubernetes, like your YAML, the metadata, metrics, and logs associated with it? In observability, we also think about real user monitoring, session replay, and really advanced analytics tools that we make available to our customers. There are some really great things about Kloudfuse that become quite different in terms of the way we're priced and the way we're deployed.

Bart: Could you share a bit about the time before and after this product announcement?

Michael: Absolutely. Before the recent announcement of Kloudfuse 3.0, we had full coverage for metrics, logs, and traces. We've recently increased the amount of sophisticated AI and ML to do anomaly detection, forecasting, and outlier detection. We've added session replay and front-end observability monitoring inside Kloudfuse. Additionally, we've added some great tools to do some really neat analysis called K-Lens, which can let us look at a heat map over time, draw a box, and discover what's different about the things that interest us and the things that are not going well.

Bart: Kloudfuse is open source and part of the CNCF landscape.

Michael: Kloudfuse itself is open source, and we use two critical Apache open source platforms inside it. Part of the infrastructure of Kloudfuse, which is deployed in a customer's Kubernetes environment - not a software as a service - we use Apache Kafka to help with data ingestion. We also use Apache Pinot as our open source backend, which is extremely performant and has awesome storage. Aside from our UI, customers can access that data directly if they need to, for example, if they wanted to run machine learning against their own observability data.

Bart: What is Kloudfuse's business model?

Michael: Kloudfuse's business model is that we're a software provider, so we build and support software. Our deployment of observability is in the customer's Kubernetes cluster. The great thing about that is there are no egress fees to send it up to a software as a service. You get to leverage your existing cloud discounts. Our product is really well optimized as far as storage. We also help you manage it. Our software runs inside your environment, making it more secure. Unlike some other solutions, we have a control plane that we run inside Kloudfuse HQ, where we can see how your cluster is doing, ensure it's healthy, and help you operate it.

Our business model is also based on the size of the cluster. Instead of having a lot of overage charges or unexpected fees, we size a cluster for a customer, provide the Helm chart to help deploy, and help manage it for them. Then, based on the amount of data, there's a flat fee for that. The nice thing about it is there's no overages. If it turns out you're sending more data to your cluster and maybe the performance isn't great, you need that data, just give us a call. We'll upgrade it, expand the cluster, and everybody's happy. No penalty, easy spend.

Bart: Who are your main competitors?

Michael: Everybody in observability is Kloudfuse's competitor. The ones that we see, though, are the prospects that come to us saying, "I'm spending too much on Datadog, New Relic, Elastic, or Splunk." Alternatively, their spend is okay, but they have fragmentation with a lot of different tools. Kloudfuse is one of the few offerings where you can address all parts of DevOps and SRE inside one product, with one user interface that is friendly and familiar to most people who use these tools.

Bart: Are there any other additional factors that help Kloudfuse differentiate from the competitors?

Michael: There are a number of factors that help us differentiate from the competitors. First, by using Apache Pinot, it becomes a data lake. With a unified data format, we can develop solutions faster, address customer demand more effectively, build things on our roadmap, and respond to customers better. We've also built in a tremendous amount of AI, which enables people to do less work to find the source of problems. Additionally, although engineers often use these tools discretely - for instance, using the log section or querying the metrics part - the APM system in Kloudfuse is so well-integrated that you almost don't have to leave the screen to figure out the source of a problem. It's really fantastic.

Bart: What should we expect next from Kloudfuse?

Michael: As we were discussing, we have a [data lake](What is the specific data lake technology used by Kloudfuse? Is it one of the mentioned backend storage solutions like Apache Pinot?) underneath Kloudfuse. What we're starting to see is that customers are asking us for help with security and other use cases, given all the observability data coming from their nodes. The fact that we have this data lake, and it's not a set of fragmented engines, makes it easy for us to prototype new solutions and enter new markets. At the same time, customers can adapt and get more value from Kloudfuse.