# Machine Learning

> An index and topic collection covering machine learning APIs, MLOps platforms, model serving infrastructure, and inference providers. Machine learning APIs span the full ML lifecycle — from data labeling, experiment tracking, and model training to model registries, hosted inference, and vector se...

This is the **Machine Learning** topic area of [API Evangelist](https://apievangelist.com) — a network of focused knowledge bases drawn from 16 years of independent API research by Kin Lane. Browse all areas at https://apievangelist.com/areas/.

## Services & Tools
- [Amazon SageMaker](https://providers.apis.io/providers/amazon-sagemaker/) (repo: https://github.com/api-evangelist/amazon-sagemaker)
- [Anyscale](https://providers.apis.io/providers/anyscale/) (repo: https://github.com/api-evangelist/anyscale)
- [Azure Databricks](https://providers.apis.io/providers/azure-databricks/) (repo: https://github.com/api-evangelist/azure-databricks)
- [Azure Machine Learning](https://providers.apis.io/providers/microsoft-azure-machine-learning/) (repo: https://github.com/api-evangelist/microsoft-azure-machine-learning)
- [Baseten](https://providers.apis.io/providers/baseten/) (repo: https://github.com/api-evangelist/baseten)
- [ClearML](https://providers.apis.io/providers/clearml/) (repo: https://github.com/api-evangelist/clearml)
- [Chroma](https://providers.apis.io/providers/chroma/) (repo: https://github.com/api-evangelist/chroma)
- [Comet](https://providers.apis.io/providers/comet-ml/) (repo: https://github.com/api-evangelist/comet-ml)
- [DagsHub](https://providers.apis.io/providers/dagshub/) (repo: https://github.com/api-evangelist/dagshub)
- [Databricks](https://providers.apis.io/providers/databricks/) (repo: https://github.com/api-evangelist/databricks)
- [Dataiku](https://providers.apis.io/providers/dataiku/) (repo: https://github.com/api-evangelist/dataiku)
- [DeepInfra](https://providers.apis.io/providers/deepinfra/) (repo: https://github.com/api-evangelist/deepinfra)
- [Determined AI](https://providers.apis.io/providers/determined-ai/) (repo: https://github.com/api-evangelist/determined-ai)
- [DVC](https://providers.apis.io/providers/dvc/) (repo: https://github.com/api-evangelist/dvc)
- [Fireworks AI](https://providers.apis.io/providers/fireworks-ai/) (repo: https://github.com/api-evangelist/fireworks-ai)
- [Google Vertex AI](https://providers.apis.io/providers/google-vertex-ai/) (repo: https://github.com/api-evangelist/google-vertex-ai)
- [Groq](https://providers.apis.io/providers/groq/) (repo: https://github.com/api-evangelist/groq)
- [Hugging Face](https://providers.apis.io/providers/hugging-face/) (repo: https://github.com/api-evangelist/hugging-face)
- [Hyperbolic](https://providers.apis.io/providers/hyperbolic/) (repo: https://github.com/api-evangelist/hyperbolic)
- [Jina AI](https://providers.apis.io/providers/jina-ai/) (repo: https://github.com/api-evangelist/jina-ai)
- [KServe](https://providers.apis.io/providers/kserve/) (repo: https://github.com/api-evangelist/kserve)
- [Kubeflow](https://providers.apis.io/providers/kubeflow/) (repo: https://github.com/api-evangelist/kubeflow)
- [Label Studio](https://providers.apis.io/providers/label-studio/) (repo: https://github.com/api-evangelist/label-studio)
- [Lepton AI](https://providers.apis.io/providers/lepton-ai/) (repo: https://github.com/api-evangelist/lepton-ai)
- [LiteLLM](https://providers.apis.io/providers/litellm/) (repo: https://github.com/api-evangelist/litellm)
- [Marqo](https://providers.apis.io/providers/marqo/) (repo: https://github.com/api-evangelist/marqo)
- [Microsoft Azure AI Foundry](https://providers.apis.io/providers/azure-ai-foundry/) (repo: https://github.com/api-evangelist/azure-ai-foundry)
- [Milvus](https://providers.apis.io/providers/milvus/) (repo: https://github.com/api-evangelist/milvus)
- [MLflow](https://providers.apis.io/providers/mlflow/) (repo: https://github.com/api-evangelist/mlflow)
- [Modal](https://providers.apis.io/providers/modal/) (repo: https://github.com/api-evangelist/modal)
- [Neptune.ai](https://providers.apis.io/providers/neptune-ai/) (repo: https://github.com/api-evangelist/neptune-ai)
- [Novita AI](https://providers.apis.io/providers/novita-ai/) (repo: https://github.com/api-evangelist/novita-ai)
- [OpenPipe](https://providers.apis.io/providers/openpipe/) (repo: https://github.com/api-evangelist/openpipe)
- [Pinecone](https://providers.apis.io/providers/pinecone/) (repo: https://github.com/api-evangelist/pinecone)
- [Portkey](https://providers.apis.io/providers/portkey/) (repo: https://github.com/api-evangelist/portkey)
- [Qdrant](https://providers.apis.io/providers/qdrant/) (repo: https://github.com/api-evangelist/qdrant)
- [Ray](https://providers.apis.io/providers/ray/) (repo: https://github.com/api-evangelist/ray)
- [Replicate](https://providers.apis.io/providers/replicate/) (repo: https://github.com/api-evangelist/replicate)
- [SiliconFlow](https://providers.apis.io/providers/siliconflow/) (repo: https://github.com/api-evangelist/siliconflow)
- [Together AI](https://providers.apis.io/providers/together-ai/) (repo: https://github.com/api-evangelist/together-ai)
- [TrueFoundry](https://providers.apis.io/providers/truefoundry/) (repo: https://github.com/api-evangelist/truefoundry)
- [Typesense](https://providers.apis.io/providers/typesense/) (repo: https://github.com/api-evangelist/typesense)
- [Vespa](https://providers.apis.io/providers/vespa/) (repo: https://github.com/api-evangelist/vespa)
- [vLLM](https://providers.apis.io/providers/vllm/) (repo: https://github.com/api-evangelist/vllm)
- [Weaviate](https://providers.apis.io/providers/weaviate/) (repo: https://github.com/api-evangelist/weaviate)
- [Weights & Biases](https://providers.apis.io/providers/weights-and-biases/) (repo: https://github.com/api-evangelist/weights-and-biases)
- [ZenML](https://providers.apis.io/providers/zenml/) (repo: https://github.com/api-evangelist/zenml)
- [Zilliz](https://providers.apis.io/providers/zilliz/) (repo: https://github.com/api-evangelist/zilliz)

## Common Features
- **Hosted Model Inference**: ML APIs from providers like Hugging Face, Replicate, Together AI, Fireworks AI, and Groq expose pre-trained and fine-tuned models behind HTTP endpoints so developers can call inference without managing GPUs.
- **Model Training and Fine-Tuning**: Platforms like Amazon SageMaker, Google Vertex AI, Azure Machine Learning, and OpenPipe expose APIs for launching training jobs, configuring hyperparameters, and fine-tuning foundation models on custom data.
- **Experiment Tracking and Model Registry**: MLflow, Weights & Biases, Comet, Neptune.ai, and ClearML provide APIs to log experiments, track metrics, compare runs, and register approved model versions for downstream deployment.
- **Vector Search and Embeddings**: Vector databases like Pinecone, Weaviate, Milvus, Qdrant, and Chroma expose APIs to index embeddings and run nearest-neighbor search powering retrieval-augmented generation and semantic search.
- **Model Serving and Deployment**: Serving frameworks like KServe, vLLM, Ray Serve, Baseten, and TrueFoundry provide APIs to deploy models as scalable HTTP or gRPC endpoints with autoscaling, batching, and routing.
- **ML Pipeline Orchestration**: Kubeflow Pipelines, ZenML, and DVC expose APIs to define, version, and execute ML pipelines spanning data preparation, training, evaluation, and deployment stages.
- **Data Labeling and Annotation**: Label Studio and similar platforms expose APIs for managing labeling projects, importing data, assigning tasks to annotators, and exporting labeled datasets for model training.
- **LLM Gateway and Routing**: LiteLLM, Portkey, and similar gateways provide unified APIs that route requests across multiple LLM providers with fallback, caching, rate limiting, and observability.

## Use Cases
- **Retrieval-Augmented Generation**: Combining a vector database (Pinecone, Weaviate, Qdrant) with an embeddings API and an LLM inference endpoint to ground model responses in private knowledge bases.
- **Fine-Tuning Foundation Models**: Using SageMaker, Vertex AI, OpenPipe, or Together AI APIs to fine-tune open foundation models on proprietary datasets and deploy the resulting model behind a managed inference endpoint.
- **Scalable Model Inference at the Edge**: Deploying optimized models through Groq, Modal, Replicate, or Baseten to serve high-throughput, low-latency inference for chatbots, recommendation systems, and content generation.
- **End-to-End MLOps Automation**: Using Kubeflow, MLflow, Weights & Biases, and ZenML to track experiments, register approved models, trigger retraining, and promote models to production via API.
- **Multimodal Application Development**: Composing image, audio, video, and text models from Hugging Face, Replicate, and Fireworks AI through standard inference APIs to build multimodal user experiences.
- **Semantic Search and Recommendations**: Indexing product catalogs, documents, or media in vector databases like Milvus or Vespa and exposing semantic search APIs to power discovery and personalization.
- **Model Observability and Cost Control**: Using gateways like Portkey and LiteLLM alongside observability platforms to monitor inference latency, cost-per-request, and routing decisions across multiple model providers.
- **Distributed Training at Scale**: Running large-scale distributed training jobs on Ray, Anyscale, Determined AI, or Databricks via API, including hyperparameter tuning and GPU cluster orchestration.

## Related Areas
- [Database](https://database.apievangelist.com): An index and topic collection covering managed databases and database-as-a-service offerings exposing APIs. Database ...
- [Migration](https://migration.apievangelist.com): An index and topic collection covering data migration, cloud migration, database migration, and API migration platfor...
- [Caching](https://caching.apievangelist.com): An index and topic collection covering API-accessible caching services, in-memory data stores, key-value caches, and ...

## More
- [Latest Machine Learning stories](/stories/)
- [All API Evangelist topic areas](https://apievangelist.com/areas/)
- [API Evangelist network index (llms.txt)](https://apievangelist.com/llms.txt)