Kloudfuse is the industry’s first unified observability platform for high cardinality data across all observability streams, seamlessly integrates into any environment, operates cost-effectively within your VPC, and delivers a smooth, SaaS-like experience.
In this conversation, CEO Krishna Yadappanavar addresses the critical paradigm shift for building adaptable, cost-efficient, and robust observability solutions when managing high data volumes and complex datasets.
Using the “Four C’s”—Cardinality, Control, Cost, and Causality Krishna is able to break down (and simplify) what’s possible with observability.
M.R.: Why has Observability gained such momentum in the last few years?
Krishna Yadappanavar: Observability has always been critical to business success, but there are a few factors that are contributing to its rising popularity.
Many organizations are dealing with overwhelming volumes of metrics, logs and traces, which can be difficult to process, let alone extract meaningful insights from. To manage these different data streams, many businesses have put in place fragmented solutions, leading to silos and inefficiencies as teams have to juggle multiple tools to discover the root cause of performance issues.
In recent years, applications have increasingly been built on microservices and cloud-native architectures. These architectures decompose applications into smaller, interconnected services, each generating its own data, resulting in an overwhelming volume of data points. The greater the volume of data, the higher the costs of observability solutions.
The growing importance of observability can be boiled down to what I call the “Five C’s”: Cardinality, Control Causality, and Consolidation, Cost.
M.R.: You talk about the five C’s, can you share more about why they are increasingly important for Observability?
Krishna: The first C is Cardinality, which reflects the challenge of handling high data volumes. As we’ve discussed, the rise of microservices as the foundation of modern application stacks has increased the amount of data that needs to be monitored. Every line of code, container it runs in, pod, cluster, user flow, service call, and database query—all contribute to performance issues. The permutations of these variables create an overwhelming number of possibilities, making it increasingly difficult to analyze and uncover insights. This poses a significant challenge for observability tools to manage such large volumes of data. Our platform is specifically designed to tackle this issue.
The second C is Control. As data becomes increasingly valuable to organizations, they are becoming more cautious about vendor lock-in. Companies want to maintain control over their own data and avoid being tied to proprietary SaaS observability platforms. Our solution offers customers greater control by enabling private virtual cloud deployments, all managed through a dedicated control plane for top-notch security and simplified administration.
Consolidation, brings together all observability data—metrics, logs, and traces, real user monitoring, continuous profiling—into a single data lake to provide a unified observability. This simplifies the process for developers by eliminating the need to switch between different tools and manually piece together different data signals to find the root cause of the problem.
The fourth C is Causality, focuses on cause-and-effect relationships in data. Troubleshooting often involves ‘unknown unknowns’ — issues that were not anticipated. While we don’t claim to pinpoint the root cause every time, as every system and problem is unique, we provide the powerful tools and insights to help you identify causal relationships more effectively and uncover deeper insights.
Finally, private deployments naturally tie into the third C, Cost. The sheer volume of the data, coupled with fragmented observability tools has led to high costs. Customers seek a more deterministic cost model that ensures predictability. Our virtual private cloud (VPC) approach and flat pricing model helps mitigate this risk by maintaining tighter control over data volumes and expenses.
Together, these five Cs—Cardinality, Control, Cost, Causality, and Consolidation—are critical to customers as they integrate Observability into their workflows to understand, monitor, and optimize the performance, health, and reliability of their systems, applications, and infrastructure.
M.R.: Kloudfuse is the second company you founded. How do you see customers behave differently in this market and at this point of time than the past?
Krishna: That’s right, there is a lot of learning that comes from having built and sold a company before. A major lesson has been the importance of product-market fit. We are laser focused on that. Most of our customers are not newcomers to observability. They have already adopted legacy tools and have experienced the growing pains that come with gen 1 products. They know the challenges well.
This familiarity makes working in an established market an advantage. Our customers know what they need, and we collaborate closely with them to ensure we meet those needs. Our rapid release cycle allows us to quickly incorporate feedback and release new features, sometimes within a two-week window.
One of the most common requests from customers is consolidation. With the fragmentation in observability tools, businesses are looking for solutions that manage both frontend and backend observability in one platform. To meet this demand, we’ve integrated real user monitoring (RUM) and session replay capabilities to complement traditional observability of metrics, logs, and traces.
Another trend we’re seeing is the demand for built-in AI features that help shorten diagnosis times and identify “unknown unknowns.” To address this, our platform includes algorithms such as Prophet, SARIMA, Pearson Correlation Coefficient, and others for anomaly detection and causality.
Cost savings are also a priority for many customers, especially those with large-scale deployments. As ARM processors become more popular for their cost-effective performance, we’ve re-engineered our platform to run efficiently on ARM instances.
Lastly, we’re seeing growing interest in generative AI and large language models (LLMs). We’re actively working with customers to integrate these capabilities into our platform.
Beyond just delivering features, we focus on understanding customer workflows, analyzing and cleansing their data, and integrating these elements seamlessly into our platform. The goal is to make workflows a first-class experience within our product. When customers adopt a platform like Kloudfuse, they don’t just get the Observability solution—it’s a comprehensive, data-driven solution that enhances our customers’ overall business performance.
M.R. Rangaswami is the Co-Founder of Sandhill.com