ade-ops takes a data platform — or the slice you own — from design to deploy, run and evolve. People and AI agents drive every operation conversationally, under the rules you set and track: fast where it's safe, gated where it counts.
One operating model from build to deploy to run to change. Currently across Databricks, Power BI and Fabric, with more platforms being added.
Humans and AI coding agents drive the same governed operations — AI-agnostic (Claude Code, Copilot, Codex), no lock-in to one assistant.
Encode naming, environments, guardrails and deploy gates once. Every operation is checked against them and logged — you always know what ran, by whom, and why.
Choose, per team and per environment, where you sit between operational flexibility and approval + role constraints — a live change can flow freely, or be shown as a diff and held for review.
Independent, composable tools for agentic data engineering — use each on its own or together. ade-catalog maps what exists and how it's connected; ade-ops orchestrates how you build, run and govern it; ade-lens consumes the certified answers people need. Map the territory, operate it — then consume it with confidence.
Analyze, explore, and document your analytics environments — a navigable catalog of what exists and how it's connected.
Operate a data platform in plain language — from design to deploy, with people and AI agents working under the rules you set and track.
Ask your certified BI estate in plain words and get the governed answer — with its provenance, in front of whoever needs it.
Live (preview) ↗A shared knowledge library for AI agents across the family.
ade-ops is open source (Apache 2.0). Start on a single project from your own machine, then scale to how your whole team operates — with people and agents in control.