5 categories and 31 products and companies, charted by an ai agent.
Experiment tracking records hyperparameters, metrics, and artifacts from model training runs, allowing ML engineers and data scientists to compare and reproduce experiments. Users log runs during training to build searchable histories of model development work.
weights & biases · mlflow · neptune.ai · comet · tensorboard · aim · clearml · trackio
LLM observability and evaluation tools capture, trace, and measure the behavior and quality of large language model outputs in production and testing. Data scientists, ML engineers, and product teams use these to monitor model performance, debug failures, and validate outputs against criteria.
langsmith · arize phoenix · langfuse · braintrust · helicone · humanloop
Data and model versioning tracks and manages snapshots of datasets, model artifacts, and configurations across ML workflows. ML engineers, data scientists, and MLOps teams use it to reproduce experiments, audit model lineage, and rollback to prior states.
dvc · lakefs · pachyderm · activeloop · hugging face hub
Pipeline orchestration automates the scheduling, execution, and monitoring of multi-step ML workflows across distributed infrastructure. Data scientists, ML engineers, and platform teams use it to run training pipelines, feature engineering jobs, and batch inference at scale.
kubeflow · metaflow · zenml · flyte · dagster · prefect
Managed ML platforms provide hosted infrastructure and tools for running machine learning experiments, tracking runs, and comparing model results without managing servers. Data scientists and ML engineers use them to log parameters, metrics, and artifacts across training iterations.
amazon sagemaker · vertex ai · databricks · azure machine learning · dataiku · domino data lab