From connected data to an answer you can check
Learn how to use Semogram and Craven to connect operational records, review AI-assisted workflows and inspect the evidence behind answers and predictions.
Get a first verified result
Start with one question and a small dataset you can inspect.
Browse by task
Getting started →
Connect a small source, save a workflow, run it and verify your first result.
Data Pipelines →
Build pipelines, manage versions, monitor runs and recover interrupted work.
Core concepts →
Understand workspaces, endpoints, ontology, evidence and prediction lifecycles.
Plugins →
Set up internal capabilities or build, test and publish a custom workspace plugin.
Data Endpoints →
Expose workspace sources, destinations and ontology stores through installed capabilities
Skills →
Create, version and evaluate reusable guidance for supported workflows.
Tutorials →
Practice bounded data workflows and measure results with data you provide.
Using your own AI assistant?
Connect a compatible client through MCP, complete the workspace authentication flow and discover the tools your permissions allow. Keep returned operation identifiers for work that continues after a call.