logistics_ops · workflow

DoorDash retrains ML demand prediction models to restore accuracy after COVID-19 disruption

COVID-19 broke the historical demand patterns DoorDash's ML models relied on, making predictions inaccurate at a time when getting supply-demand balance right had direct consequences for both customers and Dasher earnings.

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 · COVID-19 breaks demand patterns
COVID-19 restrictions caused demand to become higher and more volatile than ever before, making model retraining necessary.
Tools used
Apache AirflowDatabricksPandasLightGBM
Outcome

The retrained DPM reduced prediction variance from 7.9% to 2.8% and increased the share of predictions within the SLA percentile range from 67% to 95%, yielding a model that was much more accurate while just as precise.

What failed first

The old Demand Prediction Model began severely underpredicting demand once COVID-19 restrictions took effect. During retraining, dataset manipulation hypotheses — including downsampling pre-pandemic data and removing holiday training data — proved negligible or detrimental to model performance.

Results
Volume2.8%
Source

https://careersatdoordash.com/blog/retraining-ml-models-covid-19/

How we source this →

Grounding & classification
Source type: technical build writeup
23 fields verified against source quotes, 1 dropped as unverifiable.
forecastingpredictive analyticsfailure mode describedmetric backednamed customerproduction runtime claimedsource backedtools describedworkflow describedecommercelogisticsaccuracy improvementerror reductiontechnical build writeuplogistics opssupply chain