Skip to content

KDBL Context Lake

KDBL Context Lake (K-Lake) turns the files you already have into a searchable, governed body of knowledge — indexing what your object stores and filesystems hold and what is inside each file, so your whole estate can be searched, audited, and answered against, with grounded, verifiable AI answers on top. Your originals are never modified, and nothing leaves your environment unless you choose to send hard-to-read pages to your own cloud document service with cloud extraction.

Get started Connect your AI

Start here

  • Introduction

    What K-Lake does, the core concepts (tenants, sources, crawls, tokens), and what you'll see running.

    Read the introduction

  • Quick start

    Sign in, add a source, extract and search its content, and open the original file behind any result.

    Quick start

  • Sources

    Configure S3, Azure Blob, SMB/CIFS, SMBFS, NFS, and OneDrive/SharePoint sources — credentials, metadata caps, and lifecycle.

    Sources

  • AI over MCP

    Serve your content to Claude and other assistants over the Model Context Protocol — grounded answers, cited sources, per-user security.

    MCP server

  • Smart Actions

    Take action on your files after every crawl — extract, redact sensitive content, notify your systems — with composable per-source pipelines.

    Smart Actions

Three interchangeable interfaces

Every action is scoped to your tenant and audited. Pick whichever suits you — they act on the same data.

The operator console: dashboards, source management, content search, tokens, and a ⌘K / Ctrl+K command palette to jump anywhere.

Web console guide

kdbl-control — seed crawls, manage sources, onboard tenants, check status.

CLI reference

Programmatic access to every operation.

REST API

Versioned docs

Use the version selector in the header to read the docs for the K-Lake release you're running. latest always points at the most recent stable release.

Authoring note — embedding media

These docs support images, animated GIFs, and video. For a YouTube embed, wrap an iframe in a video-embed div for a responsive 16:9 frame:

<div class="video-embed">
  <iframe src="https://www.youtube-nocookie.com/embed/VIDEO_ID"
          title="K-Lake demo" allowfullscreen></iframe>
</div>