In your field, I’m exploring
ML platforms & training data
The answer · All companies tracked on usedby
Companies use Weights & Biases most, then H2O.ai.
58 of 309 companies tracked in this category use Weights & Biases. Every line links to a public source.
counts are companies tracked on usedby, not market shareLimited beta
- 01
58 19% - 02
53 17% - 03
49 - 04
29 - 05
28
ML platforms & training data
Tool by tool
309 companies tracked in this category · 12 tools
About these numbers
Counts cover companies recorded on usedby. They are not market shares. All four proof levels count. Missing size data is excluded only from size-based views. Your requirements have not been evaluated.
Companies with known size: 85%. Vendor companies excluded from this category: 1.

Weights & Biases
58 companies19% of the companies in this sample
37 confirmed21 observed
Meta's AI research team provides baseline learning curves in Weights & Biases for the experiments that accompany its MiniHack open-source reinforcement learning framework.
MetaRead the sourceOriginal source text
We also provide baseline learning curves in Weights & Biases for our experiments.

H2O.ai
53 companies17% of the companies in this sample
43 confirmed10 observed
Capital One's operations team uses H2O Sparkling Water and H2O-3 to forecast normal mobile app login volume with gradient boosting and flag anomalies in real time.
Capital OneRead the sourceOriginal source text
In one case, a spike of more than 20,000 unexpected logins at 11:15pm triggered an anomaly alert that reached the incident response team at 11:17pm, four minutes before any other alarm.

Hugging Face
49 companies16% of the companies in this sample
5 confirmed1 declared40 observed3 job posts
Nvidia contributes open models and datasets to Hugging Face, publishing its own models, libraries and tools in the open for developers to use and build on.
NvidiaRead the sourceOriginal source text
NVIDIA is the largest contributor of open models and data to Hugging Face, and our contributions continue to grow. NVIDIA has released more than 500 models on Hugging Face and more than 250 open datasets.

Labelbox
29 companies9% of the companies in this sample
19 confirmed10 observed
Ancestry's data science team uses Labelbox's model-assisted labeling and native image and text editors to produce labeled training data for models that extract genealogical information from historical records.
AncestryRead the sourceOriginal source text
Toward its goal of weekly model releases, Ancestry uses Labelbox as a data engine to train models faster and evaluate validation sets more easily.

Encord
28 companies9% of the companies in this sample
13 confirmed15 observed

Dataiku
28 companies9% of the companies in this sample
28 observed

Anyscale
27 companies9% of the companies in this sample
15 confirmed11 observed1 job post

DataRobot
27 companies9% of the companies in this sample
4 confirmed23 observed
Category, industry and company size
Company size describes the employer. It does not establish access for a solo practitioner.














