logistics ops
Logistics ops AI workflow patterns
Verified production AI workflows in logistics ops — including named customers, verbatim metrics, and vendor case sources. The sub-patterns below open into the common implementation shape and first-deployment failures for each.
Across 40 documented logistics ops cases
Recurring tools
lightgbm 7ocean visibility 6machine learning algorithms 4blue yonder 3docker 2grpc 2pytorch 2reform 2rpa 2sibyl 2uipath 2ai multicam 1
What fails first / common problems
The existing bulk CSV tool offered no historical evidence and forced the ETA team to query multiple databases for change history; the DTMF robocall system was limited to simple button-press responses and could not capture detailed hours …
— DoorDash 2025 Summer Intern Projects: GenAI Voice Agent, Storm Mode ETA, Probabilistic ETA Model, and LLM Alcohol RecommendationsThe manual model testing process—requiring data scientists to hand-write Python gRPC scripts per migration—was not scalable as the team grew, and early feature quality monitoring required an onboarding step that hindered adoption.
— DoorDash builds a centralized ML platform quadrupling model count and achieving 5x prediction throughputDuring the sequential swap of inference sources, two swaps produced statistically significant degradations in ASAP and DAT metrics, with degradations worse than the secondary success criterion, requiring corrective measures before contin…
— DoorDash best practices for regression-free ML model migration in Dasher assignmentWhen this work began, PyTorch had limited support for serializing computation graphs with Python dependencies.
— DoorDash ML Platform builds a computational graph system and Python DSL for flexible ensemble model production servingInitial co-training of multitask models caused significant accuracy degradation due to task interference.
— DoorDash deploys MLP-gated MoE deep learning model for 20% relative improvement in ETA prediction accuracyRepresentative reported outcomes
delays of upto 10 minutes per container · over 10,000x
Nanonets automates shipping container recognition for Adani Ports with Computer Vision and OCR
>90% · ~99%
Gelato accelerates printer and carrier onboarding via CrewAI multi-agent integration
74,000 · over 10,000
SF Supply Chain uses UiPath RPA to save 74,000 effective working hours in warehouse operations
under three minutes · close to 2,000 hours per year
Sappi deploys 132 UiPath automations saving 13,000 hours annually across European operations
24 hours' notice · circa 40 customers
Clipper Logistics uses Blue Yonder warehouse management to release inventory on 24 hours' notice and support e-commerce growth
Reported by the source case, as published — not independently verified.
Featured workflows in this category
A curated selection — highest-trust cases with the richest evidence (first-deployment failures documented, metrics on record). The full logistics ops corpus is reachable via search.
Tennis boosts warehouse efficiency 50% and cuts store replenishment cycle time 56% with Blue Yonder WMS
Blue Yonder WMS
Tennis achieved a 50% increase in warehouse efficiency, reduced store replenishment cycle time by 56% (from 4.
super.AI IDP automates Bill of Lading processing for Fortune 500 medical products company, cutting errors 95%
Super.AI IDP
super.
DoorDash 2025 Summer Intern Projects: GenAI Voice Agent, Storm Mode ETA, Probabilistic ETA Model, and LLM Alcohol Recommendations
Vapi → Kafka → Large Language Models → EntityCache
DoorDash shipped a structured Logistics Console with storm mode automation and an ETA pad auditor, a GenAI voice agent that cap….
DoorDash builds a centralized ML platform quadrupling model count and achieving 5x prediction throughput
Sibyl → Redis → gRPC
DoorDash's ML platform quadrupled the number of models and achieved 5x prediction throughput; a feature store optimization cut ….
DoorDash ML Platform builds a computational graph system and Python DSL for flexible ensemble model production serving
LightGBM → PyTorch → Sibyl Prediction Service → gRPC
The computational graph with C++ reduces CPU prediction time by more than a factor of 12 compared to Python and reduces total m….
DoorDash builds ML forecasting and optimization system to balance Dasher supply and delivery demand
LightGBM
The mobilization system reduced delivery times, cancellations, and extreme lateness for consumers; drove down merchant order ca….
Zalando trains a convolutional neural network to accelerate warehouse order batching
Caffe → cuDNN_v2 → OpenBLAS → NVIDIA Tesla K80
A convolutional neural network trained as a fast OCaPi surrogate achieved 0.
Nanonets automates shipping container recognition for Adani Ports with Computer Vision and OCR
Nanonets → OCR → Docker
Nanonets deployed an on-premise computer vision and OCR system that fully automated container data capture at port gates, reduc….