The Dagster+ AI agent uses the operational signals your team already relies on: runs, assets, failures, lineage, freshness, and system state.
- •Pipeline runs and execution history
- •Assets and asset health
- •Failures, retries, and slowdowns
- •Freshness and materialization status
- •Incidents and degraded components
See the Dagster+ AI agent diagnose pipeline issues
Ask from Dagster+ UI, Slack, or Microsoft Teams. Compass brings the operational context into the answer so the next step is clear.
Diagnose issues faster, right where work happens
Before Compass
- Things fail. Context is missing.The signal is there. The explanation isn’t.
- Noise over answersJobs fail, retry, recover. Which ones matter?
- The long way aroundLogs, dashboards, pings — still guessing.
With the Dagster+ AI agent
- Ask where the issue shows up@compass why are schedules and sensors failing in my_code_location?
- Narrow it down@compass what changed when these failures started?
- Find the root cause@compass what are the flakiest jobs this week?
Built for agentic data engineering
Works where teams investigate
Ask from Dagster+ UI, Slack, or Microsoft Teams.
Native Dagster+ context
Runs, assets, health, and history — all in scope.
Diagnosis-first answers
High-signal answers first, deeper inspection when needed.
Trusted by data-driven teams
See why teams at fast-growing companies are switching to Compass for instant, governed insights.
Speed of insights
“Compass feels different — it actually thinks like an analyst, not just another text-to-SQL tool. Now people across multiple departments use it daily. Everyone's obsessed.”
Josh
Data Scientist at Clay
Business user adoption
“Compass hit differently from the start — real analysis delivered right in Slack, exactly when and where the business needs it. My trial users are more excited about Compass than any project I've launched to date.”
Jake
Head of Data at Vanta
