Welcome to Kubling!

Kubling is a virtual-schemafull database engine that sits between clients and physical data sources.

When a data source exchanges documents, like in the case of APIs, Kubling can delegate the final processing steps (fetch and CUD operations) to custom Script modules. This provides greater flexibility by eliminating dependencies on built-in connectors.

It follows CI/CD and DevOps principles, that is, the engine loads definition and configurations dynamically from a released bundle file which, in turn, can load other dependencies as Modules.

Why?

Our goal is to apply a data-centric approach to operations, ensuring full consistency across operational platforms and systems.
By doing so, we help streamline the complexity of cloud environments, enabling organizations to regain operational control and enhance governance over their systems.

Some Use cases

• Single data source for applications: Kubling serves as an ideal sidecar for your applications, acting as a unified data source that connects your application to multiple data sources.
• AI Agents: Kubling assists your agents in discovering and interacting with hundreds of operational data sources, making the entire process more secure and easier to govern.
• Hybrid and Multi-Cloud: Manage all your cloud resources from a single platform using standard SQL in a transactional manner.
• IT Operations: Design your own abstract data model that best reflects your organization's needs, based on your current technological landscape.
• Internal Developer Platforms: Simplify the implementation of an IDP by connecting the front-end web app with all your systems through Kubling.

Is it still relevant in the AI era?

This question resonates not just for Kubling but for the entire software industry. While AI has certainly revolutionized many aspects of technology, especially with the rise of Generative Models, the reality is that AI alone cannot address every specific software need.

Generative Models are transforming how we create software, but it’s not feasible to rely on them for generating a new application from scratch every time. This isn't due to limitations in Large Language Models (LLMs), but rather because the generated software still needs to operate within predefined environments—environments governed by strict rules, particularly in enterprise settings.

While AI-driven and vector databases excel at information retrieval, when it comes to Create, Update, Delete (CUD) operations across diverse data sources, precision remains critical. Identifying specific entities, regardless of context, is essential to maintaining control and integrity in complex systems.

This is where Kubling shines. We envision lightweight, intermediate structured-language layers like Kubling playing a pivotal role in complementing LLM-based technologies. Not only does Kubling streamline operations, but its configuration (including Script Modules) can also be easily generated by LLMs, making it an adaptive and future-ready tool for modern enterprises.

What it is not?

Kubling is not designed to replace your current database system. It does not ingest or store data locally (aside from caching), nor does it offer all the features you'd expect from traditional RDBMS, NoSQL, Document, or Vector databases.

While Kubling is fully capable of handling data management and analytics tasks, it is not optimized for those purposes. Our focus is on enhancing operational data access, rather than analytical capabilities. There are many specialized Data Virtualization platforms tailored for analytics, but Kubling’s strength lies in streamlining access to operational data across diverse sources.