compliance_monitoring · logistics · workflow
DoorDash SafeChat cuts safety incidents 50% with multi-modal AI content moderation
DoorDash needed to screen a high volume of user-generated content daily — chat messages, images, and voice calls — for inappropriate material to protect Dashers and consumers at platform scale.
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 · User-generated content enters pipeline
Every piece of content exchanged between parties — text, image, or voice — enters the content moderation pipeline.
Tools used
moderation APIfast LLMprecise LLMinternal modelcomputer vision
Outcome
Since launch, SafeChat achieved a 50% decrease in low/medium-severity safety incidents and 99.8% of messages are now handled by the internal model alone, resulting in higher Dasher and consumer trust, reduced customer support volume, and stronger platform reliability.
What failed first
An initial shadow test with an external vision API for image moderation produced a significant false positive rate, making it unsuitable for production.
Results
Time savedunder 300ms
Volume50%
Grounding & classification
Source type: technical build writeup
36 fields verified against source quotes.
computer visiondocument classificationsentiment analysisvoice aicall recordingchat transcriptfailure mode describedhuman review describedmetric backednamed customerproduction runtime claimedtools describedworkflow describedlogisticsautomation ratecost reductioncustomer satisfactionerror reductiontechnical build writeupcompliance monitoringquality assuranceescalation workflowextract classify routehuman review queuemonitor detect alert