Problem
As dashboard libraries grow, users lose time browsing folders and guessing exact dashboard names.
AI dashboard search for enterprise analytics portals
Natural-language dashboard search with relevance scoring, typo tolerance, and domain-aware filtering for large BI libraries.
Built for Dodo Brands • Superset • Direction: AI & Data Analytics
AI-ranked results with match quality
Cards display a relevance score badge and a “Poor match” indicator for borderline hits, helping users quickly judge which dashboards to open. Each result is labeled with its content type (Dashboard).
Implementation narrative
Solution
We implemented natural-language search that supports fuzzy matching, synonyms, multilingual input, and visible relevance signals directly in results.
Outcome
Teams reach the right dashboard faster, reduce browsing friction, and improve daily analytics adoption.
Related features
Dodo Brands • Superset
Actionable dashboard feedback system
Open featureDodo Brands • Superset
Style Your Dashboards
Open featureDodo Brands • Superset
Organize dashboards with folders
Open featureMore views
Additional screens captured from the same implementation.
Search from anywhere
A subtle “Find dashboard with AI” input is available on the dashboard library, leveraging titles and descriptions. Results respect left-side domain filters, so users can narrow by business area before searching.
Scoring on the fly
Relevance scores are calculated instantly as the user types, handling fuzzy matches, typos, and synonyms. The UI remains minimal, emphasizing clear cards and actionable signals.