rag-services is an extensible platform that breaks down RAG pipelines into discrete microservices. It offers a document store service, a vector index service, an embedder service, multiple LLM inference services, and an orchestrator service to coordinate workflows. Each component exposes REST APIs, allowing you to mix and match databases and model providers. With Docker and Docker Compose support, you can deploy locally or in Kubernetes clusters. The framework enables scalable, fault-tolerant RAG solutions for chatbots, knowledge bases, and automated document Q&A.

SPEAR (Scalable Platform for Edge AI Real-Time) is designed to manage the full lifecycle of AI inference at the edge. Developers can define streaming pipelines that ingest sensor data, videos, or logs via connectors to Kafka, MQTT, or HTTP sources. SPEAR dynamically deploys containerized models to worker nodes, balancing loads across clusters while ensuring low-latency responses. It includes built-in model versioning, health checks, and telemetry, exposing metrics to Prometheus and Grafana. Users can apply custom transformations or alerts through a modular plugin architecture. With automated scaling and fault recovery, SPEAR delivers reliable real-time analytics for IoT, industrial automation, smart cities, and autonomous systems in heterogeneous environments.

Universal Basic Compute provides a unified environment for designing, training, and deploying AI agents across diverse workflows. Users can select from multiple large language models, configure custom memory stores for contextual awareness, and integrate third-party APIs and tools to extend functionality. The platform handles orchestration, fault tolerance, and scaling automatically, while offering dashboards for real-time monitoring and performance analytics. By abstracting infrastructure details, it empowers teams to focus on agent logic and user experience rather than backend complexity.

ToolFuzz provides a comprehensive fuzz testing framework specifically tailored for tool-using AI agents. It systematically generates randomized tool invocation sequences, malformed API inputs, and unexpected parameter combinations to stress-test the agent’s tool-calling modules. Users can define custom fuzz strategies using a modular plugin interface, integrate third-party tools or APIs, and adjust mutation rules to target specific failure modes. The framework collects execution traces, measures code coverage for each component, and highlights unhandled exceptions or logic flaws. With built-in result aggregation and reporting, ToolFuzz accelerates the identification of edge cases, regression issues, and security vulnerabilities, ultimately strengthening the robustness and reliability of AI-driven workflows.

The Agent Supervisor Example repository demonstrates how to orchestrate several autonomous AI agents in a coordinated workflow. Built in Python, it defines a Supervisor class to dispatch tasks, monitor agent status, handle failures, and aggregate responses. You can extend base agent classes, plug in different model APIs, and configure scheduling policies. It logs activities for auditing, supports parallel execution, and offers a modular design for easy customization and integration into larger AI systems.

AgentMesh provides a modular infrastructure for developers to create networks of AI agents, each focusing on a specific task or domain. Agents can be dynamically discovered and registered at runtime, exchange messages asynchronously, and follow configurable routing rules. The framework handles retries, fallbacks, and error recovery, allowing multi-agent pipelines for data processing, decision support, or conversational use cases. It integrates easily with existing LLMs and custom models via a simple plugin interface.