Reading
OhWise is a multi-agent AI platform that enables intelligent coordination across data, documents, and workflows through DAG orchestration, AI coding agent lab, Studio coordination pipelines, and graph-native context retrieval.
Section 01
OhWise is a multi-agent AI platform that enables intelligent coordination across data, documents, and workflows through DAG orchestration, AI coding agent lab, Studio coordination pipelines, and graph-native context retrieval.
Section 02
Section 03
With the explosive growth of large language models and AI applications, enterprises and developers are facing a new pain point: tool fragmentation. Data is stored in vector databases, document management uses RAG systems, code generation relies on Copilot-like tools, and workflow automation uses another set of systems—users have to switch between dozens of tools, and the problem of data silos is becoming increasingly severe.
The OhWise project was born to address this pain point. It is a multi-agent AI platform whose core concept is "coordination"—allowing multiple specialized AI agents to work collaboratively, connecting data, documents, and workflows, so users can get complete intelligent services without having to piece together various tools.
Section 04
OhWise provides five core capability modules:
DAG-based Orchestration: Uses Directed Acyclic Graph (DAG) to define and orchestrate complex multi-agent workflows. This visual orchestration method allows developers to intuitively design the dependencies and execution order between agents.
AI Coding Agent Lab: A dedicated environment for AI-assisted programming. Developers can interact with AI coding agents here to get services like code generation, code review, and refactoring suggestions.
Studio Coordinator Pipelines: A pipeline system for coordinating the collaboration of multiple agents. It is responsible for agent scheduling, state management, and result aggregation.
Graph-Native Context Retrieval: Unlike traditional vector retrieval, OhWise uses knowledge graphs as the core mechanism for context retrieval. This method can capture complex relationships between entities and provide more accurate context understanding.
Enterprise Multi-Tenancy: Provides complete tenant isolation, permission management, and resource quota control for organizational-level deployments.
Section 05
ohwise-web-hub is the official website repository of OhWise, using the current mainstream front-end tech stack:
Vite: As a build tool, it provides an extremely fast development experience and optimized production builds.
React + TypeScript: Type-safe component-based development, improving code quality and maintainability.
Tailwind CSS: An atomic CSS framework for rapid style development.
shadcn/ui: A high-quality React component library based on Radix UI, providing a consistent design language.
This technology selection reflects the project's dual focus on development experience and user experience.
Section 06
OhWise has built a complete multi-agent open source ecosystem, including five core projects:
ai-relay: A WebSocket relay service that provides real-time communication capabilities for the AI coding agent CLI. This is the infrastructure for the Lab function.
ohwise-mcp: MCP (Model Context Protocol) server, providing graph toolset and Studio pipeline integration capabilities.
graph2sql: An open-source Text-to-SQL tool that combines Schema graphs and PPR (Prompt Plan Representation) technology to convert natural language queries into precise SQL statements.
docs2graph: A document knowledge graph extraction tool that can automatically build knowledge graphs from documents like PDF and Word.
codebase2graph: A codebase knowledge graph extraction tool that analyzes source code repositories and builds graph representations of code entities and dependencies.
These five projects work together to form a complete chain from data ingestion (docs2graph, codebase2graph) to intelligent querying (graph2sql) and then to agent coordination (ohwise-mcp, ai-relay).
Section 07
The OhWise platform is suitable for various enterprise-level AI application scenarios:
Enterprise Knowledge Management: Integrate documents, codebases, and databases scattered across departments to build a unified enterprise knowledge graph. Employees can get precise answers through natural language queries.
Intelligent Customer Service: Multiple agents collaboratively handle customer inquiries—one agent is responsible for intent recognition, one for knowledge base retrieval, one for response generation, and the Studio coordination pipeline ensures their seamless collaboration.
Code Intelligence: AI coding agents can not only generate code but also understand the architecture and dependencies of the entire codebase, providing context-aware code suggestions.
Data Analysis: Through graph2sql, business personnel can query complex databases with natural language without mastering SQL syntax.
Process Automation: The DAG orchestration capability allows enterprises to visually design complex business processes, and agents are responsible for executing specific tasks.
Section 08
The design philosophy of OhWise can be seen from the project structure:
Modular Architecture: Each function (Lab, Studio, knowledge graph) is an independent module that can be used alone or in combination.
Open Standards: Uses open protocols like MCP to ensure interoperability with other AI tools and platforms.
Progressive Adoption: Enterprises can start with a single module (e.g., using docs2graph for document knowledgeization) and gradually expand to the complete platform.
Developer-Friendly: Complete TypeScript type definitions, clear API documentation, and rich example code reduce the entry barrier.
Keep going with more reads from the same topic.
Nornir MCP Server is an enterprise-level server based on the Model Context Protocol (MCP). It seamlessly integrates large language models (such as Claude) with the Nornir network automation framework, supporting natural language orchestration for multi-vendor network devices (Cisco, Arista, Juniper, etc.), and providing production-grade features like a dual-engine architecture (NAPALM + Netmiko), intelligent filtering, and a secure sandbox.
Bibliothèque Française LLM is a structured indexing and annotation project for French public domain literature designed specifically for large language models (LLMs). It integrates multiple authoritative sources such as DraCor, Common Corpus, and Wikisource, providing metadata indexing categorized by genre, author, and era, as well as in-depth annotations for dramatic texts (including characters, lines, stage directions, etc.). Its aim is to enable LLMs to efficiently read and understand classic French literary works.
Splinter is a minimalist, high-performance key-value (KV) and vector storage system enabling zero-latency inter-process communication via shared memory and atomic operations. With only 766 lines of core code, it supports millions of operations per second and 768-dimensional vector storage, offering a new architectural approach for local LLM inference and data-intensive applications.
libmlxforge is an embeddable MLX large language model (LLM) inference engine designed specifically for Apple Silicon. It provides a unified C ABI interface, supports calls from Node.js, Swift, and Rust, and features continuous batching, streaming output, JSON-constrained structured output, and embedding vector generation.