Unilever · Google Cloud
15-35% increase — retailer sales
Independently reportedOther · Other · Classical ML · Scaled
Other
Rabot · Lightly
Vendor-reportedManufacturingUnited StatesScaledMid-marketComputer vision
Vendor-reported. The customer is named and the numbers are quoted from the source page, but the account comes from the vendor. No independent confirmation.
Reported by lightly.ai (vendor self-report). Checked against the source page on 2026-09-06. 4 of 4 figures below appear on that page word for word.
The problem
Rabot's customers have different packing station layouts, requiring model retraining for each. This led to scalability issues and inefficient data selection for training.
What was deployed
- Vendor
- Lightly
- Products
- Lightly platform, LightlyTrain
- Technique
- Computer vision
- Build or buy
- Bought off the shelf
- Deployment
- SaaS
- Data used
- packing-station frames from millions of videos, images/videos, 50M images
- Scale
- process over 50M images within 6 months
What changed
Each row is quoted from the source. Figures we could not find on the page in those words are marked — they are kept, not deleted, so you can judge them.
2x fastercustomer onboarding speed
“enabling Rabot to 2x its onboarding speed”
Quoted word for word from the source · round percentage, no baseline given
retraining process time
“50% Reduced Retraining Process Time”
Quoted word for word from the source · round percentage, no baseline given
over 50Mimages processedwithin 6 months
“process over 50M images within 6 months using Lightly.”
Quoted word for word from the source
model accuracy
“improve their model accuracy by 10%.”
Quoted word for word from the source · round percentage, no baseline given
“Lightly is hyper-focused on finding thousands of relevant images from millions of video frames to improve deep learning models. The Lightly platform enabled us to build models and deploy features more than 2x faster and unlock completely new development workflows.”
Difficulties and limits
- Scalability: Retraining the model requires a lot of data from the customers. Those images/videos then need to be analyzed to determine which images should be annotated and used for retraining.
- Selection: Selecting the right data for training is essential for seamless integration. Existing selection processes were neither efficient nor could they provide an intelligent data selection.
Sources
- lightly.ai — vendor self-report
Similar deployments
ANZ Bank · NVIDIA
0.82 — Gini coefficient for assessing risk
Independently reportedBanking & finance · Other · Classical ML · Pilot
BHP Billiton
zero — accidents due to driver drowsiness
Independently reportedOther · Other · Classical ML · In production · 2022
DSST: College View · Playlab, Class Companion, Manta AI
Vendor-reportedEducation · Other · NLP · In production · 2024
Cubo Ai · Google Cloud
more than 10X — user growth supported with same IT workforce
Vendor-reportedOther · Other · Computer vision · Scaled · 2019
Precision HVAC & Cooling · AutoScale My Biz
15+ — High-Value After-Hours Leads Captured
Vendor-reportedOther · Other · LLM · In production