
A place for AI forward engineers and leaders
Watch sessions on building AI agents grounded in your governed data.


## The problemA single analytics engineer often serves hundreds of downstream consumers. Every why-is-this-number-wrong conversation costs hours and damages trust. Worse, model changes ship without anyone enumerating the downstream datasets, dashboards, and report owners that will quietly break. Health drift goes undetected until a stakeholder notices.## What this app doesTwo agents share one workspace. The lineage scanner walks the full dataset-to-dashboard graph for any selected upstream model and returns a complete impact map — every downstream dataset, dataflow, card, and dashboard in the chain — then dispatches a plain-language change notification to the identified owners. The health monitor queries each registered core model on a schedule for row-count trends, null rates, schema fingerprint, and distribution against baseline, then surfaces flagged models in a morning digest. Every check writes to a running audit log.## Key takeaways- Full lineage map before any model change ships- Targeted owner notifications, not mass-blast- Daily health drift catches issues before stakeholders do- Living audit trail for governance and trust## Who it's forAnalytics engineering teams, data platform owners, and BI leads supporting hundreds of downstream consumers off shared models.