Quality assurance · Production

Dialpad achieves 20% labeling quality improvement and 41% cost reduction with Labelbox

The problem

Dialpad's legacy crowdsourced annotation provider delivered inaccurate and missing labels that fell short of quality standards, causing data scientists to grow hesitant to request training data at all because of the time and effort involved.

First attempt

The previous labeling provider used crowdsourced labeling whose quality steadily declined; extra time and resources spent redesigning labeling projects still could not meet quality requirements.

Workflow diagram · grounded in source
1
Training data need identified
trigger
“Building and maintaining these models require large amounts of training data for custom models as well as subject matter expertise and fine-tuning when it comes to LLMs”
2
Submit labeling project to Labelbox
integration
“The team settled on Labelbox as their new labeling solution for handling a variety of NLP and LLM-focused tasks”
3
Boost assigns best labeling team
routing
“Dialpad also leveraged Labelbox Boost to find and employ the best labeling team for their use cases, reducing the need for extra supervision from Dialpad's data scientists”
4
Labeled training data delivered
output
“produce higher-quality labeled data for their AI use cases”
Reported outcome

After a year with Labelbox, Dialpad saw a 20% improvement in labeling quality and a 41% reduction in labeling costs, with data scientists now proactively requesting training data and able to scale AI development faster.

Reported metrics
Labeling quality improvement20%
Labeling cost reduction41%
Cost per data point (previous provider)roughly 29 cents per data point
cost per data point (with Labelbox)15 cents per data point
Show all 7 reported metrics
labeling quality improvement20%
labeling cost reduction41%
cost per data point (previous provider)roughly 29 cents per data point
cost per data point (with Labelbox)15 cents per data point
cost per data point reduction percentage48%
accuracy increaseover 20%
team productivity and motivationmarked improvement in team productivity and motivation
Reported stack
LabelboxLabelbox Boost
Source
https://labelbox.com/customers/dialpad-customer-story/
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

After a year with Labelbox, Dialpad saw a 20% improvement in labeling quality and a 41% reduction in labeling costs, with data scientists now proactively requesting training data and able to scale AI development faster.

What tools did this team use?

Labelbox, Labelbox Boost.

What results were reported?

Labeling quality improvement: 20%; Labeling cost reduction: 41%; Cost per data point (previous provider): roughly 29 cents per data point; cost per data point (with Labelbox): 15 cents per data point (source-reported, not independently verified).

What failed first in this deployment?

The previous labeling provider used crowdsourced labeling whose quality steadily declined; extra time and resources spent redesigning labeling projects still could not meet quality requirements.

How is this quality assurance AI workflow structured?

Training data need identified → Submit labeling project to Labelbox → Boost assigns best labeling team → Labeled training data delivered.

Related quality assurance cases