compliance_monitoring · workflow
DoorDash dark shipping: safely iterating ML fraud detection models in production
DoorDash's anti-fraud ML models must meet stringent reliability and correctness requirements while needing rapid iteration as fraud patterns evolve — a tension that pre-production testing alone cannot resolve because correctness can only be fully verified with real production traffic.
How it works
Common implementation structure
How this type of workflow is generally built, generalized across documented cases — not tied to any one vendor's stack. Click any stage to read what happens there. Specific products that implement these stages appear in “Tools commonly seen” below.
Stage 1 · Pre-production iteration
The model is iterated rapidly in development environments — updated, trained, evaluated, and tuned — until backtesting results look consistently good.
Tools used
machine learning platformrule engineML serviceCurie experimentation system
Outcome
The dark rollout practice enables DoorDash to iterate on production ML deployments quickly while minimizing the risk of regressions, by validating models with real production traffic before they make live decisions.
Results
Volumemany millions of times each day
Grounding & classification
Source type: technical build writeup
17 fields verified against source quotes.
fraud detectionpredictive analyticsfailure mode describednamed customerproduction runtime claimedtools describedworkflow describedlogisticscycle time reductionerror reductiontechnical build writeupcompliance monitoringmonitor detect alert