DoorDash, in partnership with Observe.AI and Amazon Web Services, has automated quality evaluations for nearly 100% of customer interactions across its 19,000-agent support operation. The move shifts the company's quality assurance from manual sampling of a fraction of calls to a system that surfaces real-time behavioral insights and accelerates detection of product friction points.
From Manual Checkbox QA to Diagnostic Insight
As DoorDash expanded globally, traditional QA processes-based on limited sampling and binary scoring-could no longer keep pace. The company needed a more diagnostic approach that could uncover the behavioral drivers behind customer pain points, not just box-ticking compliance metrics. Through Observe.AI's platform, powered in part by AWS transcription and AI infrastructure, DoorDash built a conversational intelligence engine that evaluates almost every interaction automatically.
"We didn't want automation for its own sake," said Xenia Strunnikova, Head of Customer Experience, Fraud, Trust & Safety S&O at DoorDash. "We wanted to better understand what drives customer sentiment, move beyond binary scoring, and put the customer at the center of our decisions. This partnership enabled us to do that at scale."
Automation freed human quality teams from manual scoring so they could focus on higher-value analysis-identifying subjective issues, interpreting behavioral signals, and ensuring customer safety and fairness across interactions.
Real-Time Hotspot Detection
Observe.AI's conversational intelligence signals-including sentiment, comprehension metrics, and behavioral indicators-allow DoorDash to infer customer satisfaction even when no explicit feedback is given. The impact on issue detection is measurable. Emerging product friction points, such as DashPass concerns or problems with new feature launches, that once took days to uncover can now be surfaced in near real time.
"With nearly 100% coverage, we can proactively identify hotspots we might never have seen before," said Joaquin Dufeu, Director of Strategy & Operations for Customer Experience & Integrity (CXI) at DoorDash. "What used to take days now happens almost immediately. That changes how we operate-and how quickly we can improve the experience."
Building a Platform Through Co-Creation
The companies describe the work as an iterative co-creation, not a simple tool deployment. Observe.AI and DoorDash built and refined signals together in cycles, ensuring the platform reflected DoorDash's operational realities. "This was about building a platform alongside DoorDash that enables automation, diagnostic insight, and real-time intelligence at enterprise scale," said Deepak Kumar, Observe.AI's Chief Customer Officer. The ongoing collaboration intends to deepen real-time capabilities and expand the range of signals used to inform customer experience decisions.
Why This Matters for Customer Support Teams
For customer support professionals, this initiative signals a clear shift in what AI means for QA and operations. Instead of replacing human judgment, the system supplies objective, large-scale behavioral data that augments how teams coach, prioritize, and respond. The core skill moving forward is not just reviewing interactions but interpreting the diagnostic signals that conversational intelligence provides. As organizations adopt AI for Customer Support, the ability to move from sample-based scoring to full-coverage analysis will reshape how teams measure quality, train agents, and loop customer feedback into product development. The DoorDash-Observe.AI approach shows that automation can turn QA from a backward-looking audit into a real-time operational intelligence function.
Your membership also unlocks:
