AI Infrastructure & MLOps

Kedro-Viz visualizes Kedro data and machine learning pipelines and tracks experiments.

BaseAI helps developers build serverless AI agents with memory, tools, and composable workflows for local development and deployment.

FastChat offers an open platform for training, serving, and evaluating large language models, and hosts release repositories for Vicuna and Chatbot Arena.

Burn provides a flexible, efficient, and portable tensor library for building deep learning applications.

Apache SINGA supports distributed deep learning through a system and library for model training.

T-Rex Label uses visual models for image annotation, including bounding boxes and segmentation, and supports COCO and YOLO formats.

MiroShark describes a swarm intelligence engine for simulating scenarios with stated time and cost targets.

Flexible, high-performance serving system for machine learning models.

Harbor sets up a preconfigured LLM stack with one command and includes hundreds of services to explore.

Mirascope describes itself as an alternative approach to building applications with large language models.

BurnRate tracks AI model provider spending in a dashboard and offers a local CLI for agent-level costs.

Pathway offers cloud templates for RAG, AI pipelines, and enterprise search that stay synchronized with connected data sources.

The fenn Python framework supports machine learning workflows and LLM agents, with trainers, templates, logging, and configuration management.

ModelFox supports training, deployment, and monitoring for machine learning models.

Cascade supports MLOps workflows for smaller projects without a full platform.

Nos dynamically partitions GPU resources and manages quotas in Kubernetes clusters.

Instill Core coordinates data, models, and pipelines for AI first applications.

Deep learning in Rust, with shape checked tensors and neural networks.

Helicone provides an AI gateway for routing and monitoring AI applications.

Eden AI routes requests across language and other AI models through a unified API with provider options.

Scalable, fast, and disk-friendly vector search in Postgres, the successor of pgvecto.rs.

The cONNXr project provides a pure C runtime for ONNX models intended for embedded devices.

E2B provides an open-source secure environment with tools for enterprise agents.

DataHub describes itself as a context platform for data and AI stacks.