AI Infrastructure

Latitude provides observability for AI agents and helps teams trace failures and verify fixes.

MMDeploy provides a model deployment framework from the OpenMMLab project.

Labelbox provides data and simulated environments for training AI systems and deploying agents across business applications.

Talos provides tools for running hyperparameter experiments with TensorFlow and Keras.

Quilt manages scientific data on AWS with versioned data packages for teams and AI.

Mem0 provides a persistent memory layer that helps AI agents and applications learn from past user interactions.

Deploy a ML inference service on a budget in less than 10 lines of code.

The MegEngine framework supports deep learning and automatic differentiation, according to its repository description.

Use Aeromancy to build reproducible machine learning workflows and projects.

Scenario supports custom workflows for image, video, audio, and 3D, with style training and tools for developers.

OpenModelZ scales LLM inference on Kubernetes with engines including vLLM, SGLang, and LMDeploy.

Zep unifies business data, documents and conversations into governed context for enterprise AI agents.

N2D2 provides an open source CAD framework for deep neural network simulation and application building.

Arize Phoenix provides observability and evaluation tools for AI applications and systems.

Chunk Tuner benchmarks document chunking strategies for RAG and recommends a configuration through a CLI, Python library, or MCP server.

VoiceAIWrapper lets agencies offer white-label AI voice agents through providers including Vapi, Retell, ElevenLabs Agents, Bolna and Ultravox.

Sagify describes its project as a way to work with large language models and machine learning more easily.

The open source Feast feature store lets teams define, manage, discover, and serve features for machine learning.

ORMB packages machine learning models as OCI artifacts, in a workflow described as Docker for ML models.

Cog packages machine learning models into containers to support deployment across environments.

Compose, test, and optimize prompts for language-model applications in Promptmetheus, an IDE supporting more than 100 models.

Rockpool supports training spiking neural networks with PyTorch and JAX and deploying them to neuromorphic hardware.

Deepchecks tests and validates machine learning models and data, covering workflows from research through production.

Neurolink connects AI model providers to applications and includes generation, streaming, voice, RAG, and file processing features.