Databricks Genie Ontology — AI skill for Claude Code
A working Databricks Genie Ontology setup (metric views, domains, pages, certification) on a retail star schema — tested live against Genie One and a Genie Agent, with research on Power BI and AI-agen.
How to install Databricks Genie Ontology
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open ParthGulavani7/databricks-genie-ontology and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Databricks Genie Ontology does
A working Databricks Genie Ontology setup (metric views, domains, pages, certification) on a retail star schema — tested live against Genie One and a Genie Agent, with research on Power BI and AI-agent (Claude/MCP) access.
Alternatives in AI
- Guizang Ppt Skill — AI-agent Skill for generating polished HTML slide decks: editorial magazine and Swiss layouts, image prompts 24.9k ★
- Zen MCP Server — The power of Claude Code + Gemini / OpenAI / Grok / OpenRouter / Ollama / Custom Model working as one 11.3k ★
- Geo SEO Claude — GEO-first SEO skill for Claude Code 3.6k ★
README
Genie Ontology on Databricks: Setup, Research & Findings
A working Databricks **Genie Ontology** setup (Unity Catalog semantics: metric views, domains, pages, certification) built on a small "Northwind Retail" star schema, plus research into how it compares and connects to Power BI and external AI agents like Claude.
What's in here
- `docs/` — the write-ups: what Genie Ontology is, how this setup was built step by step, live tests against the Genie Agent and Genie One, and four research findings comparing it to Power BI and to AI-agent access patterns.
- `setup/` — the actual notebooks used to build the setup in Databricks
(catalog/schema/tables, metric views, Genie Agent, domains, certification, pages).
Credit: adapted from the public demo repo
databeli/genie_ontology, run against
our own workspace (catalog
genie_ontology_demo, schemaretail_demo). - `screenshots/` — screenshots of the live setup: the schema, a metric view and its measure definitions, business domains, a business definition page, the Genie Agent's configuration and test run, and a live Genie One query.
- `final-docs/` — the polished deliverables: a slide deck
summarizing the setup and research findings for a team walkthrough
(
genie-ontology-deck.pptx), and a captioned Word doc of the setup screenshots (genie-ontology-setup-screenshots-v4.docx).
Quick summary
Genie Ontology is Unity Catalog's governed semantic layer for Databricks Genie: it gives Genie (and any AI agent reading from it) the business context raw tables don't carry on their own; definitions, formulas, groupings, and trust signals, so answers are accurate and consistent instead of guessed. This project builds that layer for a small retail dataset (4 dimension tables, 3 fact tables, 3 metric views, 4 business domains, 3 definition pages, one Genie Agent), tests it live against both th
Related Skills
Psql AI Assisted Development
Reference implementation for AI-assisted PostgreSQL development: star-schema KPI views, incremental materialis
Rails AI Context
45 MCP tools that give AI coding agents ground truth about your Rails app: schema, models, routes, controllers
Karpathy Wiki Protocol
Battle-tested protocol for AI agents (Claude Code, Codex, Cursor) to build and maintain Karpathy-style wikis.
Pbi Status
Show what is open in Power BI Desktop (Bridge status, unsaved changes, pages), which local model instances the
Agentic Customer Support
Production-grade AI-powered customer support Agent System built on Databricks.
Wow Addon Dev
WoW addon-development skill for AI coding agents (Claude Code, Codex, Gemini CLI). Retail Midnight 12.0+ and C
Related Agents
North Star
Define and maintain the North Star metric and cascade OKRs. Use when the team needs a single measurable outcom
Domain Extractor
Layer-0 subagent for the rev-3 agentic-ontology methodology. Reads the target project's live documentation (We
Pbi Reviewer
Read-only quality gate for a Power BI project: checks the live model (structure, DAX hygiene, date table, RLS)