TL;DR:
The combo of Codex (AI codegen/review) and Cursor (AI IDE/context) is now my default stack for any dev work—not just PowerBI. The key: plugging both into our Planner (tasks) and Cortex (docs/decisions) so every step is tracked, reviewable, and shareable.
For those not in our stack:
- Planner = advanced project/task tracker (dependencies, AI action logs).
- Cortex = semantic doc/knowledge base (like Notion/Confluence, but with built-in AI for tagging, context, and audit).
- Codex = code reviewer, optimizer, and PR fixer (via cloud or in-editor).
- Cursor = IDE/workspace, doc/test generator, and context anchor.
PowerBI example (real use):
- Built and tested a PowerBI Dataflow JSON Validator in Cursor,
- Used Codex for code review/PR feedback,
- CI/CD ensures every change goes through the same process—no shortcuts, all traceable.
But it’s not just PowerBI:
- I’m using this same Codex+Cursor + Planner+Cortex combo for new app builds, backend services, even internal tooling.
- The benefit: all process, logic, and decisions are linked together—you can trace any feature or fix from code, to doc, to planner task, to review history.
Why it matters:
- Onboarding = instant (all links and context in one place).
- Review/governance = built-in, not bolted on after the fact.
- Works for any coding project, not just dashboards.
If you’re building apps, pipelines, or internal tools and want AI and governance, try connecting your copilot/IDE to your project and doc stack.
DM if you want examples or want to compare notes!