Organizations have users tables which store information about users belonging to the organization.

DataLens consults the UX Expert when needing validation or feedback on user experience concerns.

The Discovery Service applies the Value overlap join strategy to identify joins by data overlap with 75% confidence.

Vanna 2.0 is a comparable AI data platform to DataLens, with differences including multi-LLM and multi-DB support, whereas DataLens currently has single LLM and single DB focus. The architecture comparison document validates design decisions related to Vanna 2.0. Vanna 2.0 uses user-aware identity propagation through system prompts, tool execution, and SQL filtering. Vanna 2.0 uses a tool registry to extend tools with access group permissions. Vanna 2.0 provides streaming UI components to stream structured response objects like tables and charts. Vanna 2.0 utilizes lifecycle hooks for quota checking, logging, and content filtering at request lifecycle points. Vanna 2.0 provides a pre-built web component for embedding the chat interface in various frontend frameworks. Vanna 2.0 enforces row-level security by filtering queries per user permissions. Vanna 2.0 tracks every query per user for compliance through audit logs. Vanna 2.0 architecture is governed by user-aware identity flow at every layer Vanna 2.0 architecture uses a tool registry for extensible tools with permissions Vanna 2.0 architecture supports streaming UI components with progress updates and structured data Vanna 2.0 architecture uses lifecycle hooks for quota checking, logging, and content filtering Vanna 2.0 enforces row-level security filtering queries per user permissions Vanna 2.0 records audit logs tracking queries per user for compliance

Requirement to monitor vector embedding progress; recently fixed for accurate reporting.

Vectorize Progress Tracking uses the RQ job queue to track status and progress of asynchronous embedding jobs. Vectorize Progress Tracking obtains chunk counts from DuckDB service to compute embedding progress accurately. Vectorize Progress Tracking queries the RQ job queue and chunk counts to provide accurate vectorization progress percentages. Direct communicator Jesper prefers the longer fix implementing vectorize progress tracking rather than just quick AI summary fixes.

Background workers include the vectorize worker as a component. The batch processor orchestrator depends on the vectorize worker to generate embeddings for processed chunks.

VisualizationService uses VisualizeSkill to generate automatic data visualizations using Lux.

DataLens Agent Mode includes the visualize skill for generating interactive charts using Plotly.

VisualizeSkill uses SkillResult when generating Plotly chart specifications. VisualizationService uses VisualizeSkill to generate automatic data visualizations using Lux.

Word count validation constrains the scope field to have a hard minimum of 20 words. ProjectCreate validation for scope requires word count validation to enforce text length.

WorkflowService uses BatchProcessor to orchestrate analysis workflows and data processing pipelines. WorkflowService orchestrates pipelines that may be started and monitored by BatchProcessor.