Datasources
SnowflakeDatabricksBigQueryPostgres+15
THE PLATFORM
A Unified Semantic
Context Graph
Define once. Deliver it in context everywhere across AI.
THE ARCHITECTURE
Three primitives. One semantic substrate.
Kanlet is an autonomous semantic context layer. In practice, this means it brings metadata graph, decision memory and semantic knowledge together into a single context structure so AI agents are reasoning from the trusted source of truth. Underneath that, Kanlet runs on three core primitives.
01STRUCTURE LAYER
Metadata Graph
Mapping every data source into a single, queryable graph, connecting tables, columns, glossary terms, dashboards and policies into one machine-readable layer.
stored as
concepthierarchymetadatalineage
02KNOWLEDGE LAYER
Memory & tribal knowledge
Capturing the tribal knowledge through decision traces - what was decided, who decided it, and why - so the next agent inherits what the last one learned.
stored as
memory notesaudit logsfeedback
03MEANING LAYER
Semantic knowledge
Encoding ontology, glossary, terms, and lineage directly into a knowledge graph, giving every query the organizational context needed to interpret data correctly.
stored as
entityedgerelationproperty
HOW WE BUILD
From buried meaning to trusted context.
From capturing institutional meaning to shipping production-ready semantic context, Kanlet guides you through every step of semantic model definition and refinement.
Step 01
Capture shared meaning
Business meaning is distributed across systems, reports, documentation, workflows, and the people who use them. Kanlet brings this organizational knowledge together into one semantic graph.
Step 02
Encode it in a semantic model
Kanlet encodes shared business meaning into a governed semantic model that can be refined continuously. Teams can update definitions, relationships, and rules as new questions, decisions, and real-world exceptions reveal where the model needs to evolve.
Step 03
Govern, secure & refine
Built-in governance and permissioning tools let your data teams classify, protect, and manage access to every data asset with full audit trails. Role-level controls, data classifications (PII, PHI, SPII), and inline review workflows ensure what ships is exactly what your team trusts.
Step 04
Deliver it everywhere it's needed
Your data and its meaning only matter if they reach the moments decisions get made. Kanlet makes your semantic model production-ready quickly. Access your semantic context through Web-UI, MCP server, and REST APIs.
THE ENGINE
Living Context Graph
Capture semantic information from your warehouses, data platforms, BI systems, and documentation.
Refresh the context graph as source systems evolve.
Data catalogs
RightDataSecodaAtlasDataplex
BI & transformation
Power BIdbtTableauLooker
Documentation
ConfluenceNotionGoogle DriveSharePoint
Regularly synced as schemas, terms, and sources change
READY WHEN YOU ARE
Close the Context Gap.
Make AI Work in Production.
See how Kanlet connects fragmented data, business meaning,
and tribal knowledge into a shared semantic layer for AI agents.