Wenjie Zhang
Wenjie Zhang is an individual software-engineering Ph.D. student and solo developer/researcher, not an incorporated company. His work focuses on NLP, knowledge graphs, entity-relation extraction and LLM applications; his open-source Yuxi project aims to put team knowledge, tools and multi-agent execution into a privately deployable workspace.
At a Glance
- Enterprise and technical teams needing private deployment
- Individual developers and open-source contributors
- Universities and research teams
- Organizations working with internal documents, knowledge graphs and agent workflows
AI Tools by Wenjie Zhang
(1)
Yuxi
Self Hosted Knowledge Agent Platform
Discussions
No discussions yet
Be the first to start a discussion about Wenjie Zhang
Latest News
Yuxi documentation reorganized and outdated/duplicate material corrected; new intro, deployment and model-usage documentation added.
Added direct multi-image chat input, including paste/drag/drop handling, image limits and history/run preservation.
Added per-sandbox resource limits and Milvus log rotation for deployment operations.
Yuxi v0.7.2 beta introduced migration handling for workdirs, system configuration, shared Skills and runtime identities.
Products & Services
MIT-licensed, self-hosted, multi-tenant knowledge-agent platform. It combines a knowledge base and RAG, knowledge graphs, LangGraph multi-agent orchestration, MCP/Skills, sandbox tools, file artifacts, API keys and team permissions in one workspace. Deploys with Docker Compose.
Earlier/open-source knowledge-graph construction and reasoning / knowledge-base question-answering system listed on Zhang's official site and GitHub projects.
Prompt-based learning method for relation classification using relation-label tokens, semantic initialization, entity-aware contrastive learning and an attention-query strategy.
Fast framework for entity-relation extraction, listed by Zhang as accepted by Data Mining and Knowledge Discovery.
Market Position
Yuxi is positioned as a self-hosted, multi-tenant knowledge-agent workspace rather than only a chat UI or a vector database: it unifies RAG, knowledge graphs, multi-agent orchestration, extensibility and governance. Relevant alternatives include Dify and RAGFlow for knowledge/RAG applications, and LangGraph-based custom agent stacks; Yuxi differentiates through an integrated private-deployment boundary and built-in team permissions, sandboxing and artifacts.
Leadership
Founders
Wenjie Zhang (张文杰)
Software Engineering Ph.D. student at Jiangnan University from fall 2023; GitHub identifies him as a Jiangnan University Ph.D. student in Computer Science and Technology (class of 2023). His site lists research in NLP, knowledge graphs and entity-relation extraction. He received a B.S. in Computer Science and Technology from Jiangnan University in 2021, according to the LabelPrompt paper biography.
Executive Team
Wenjie Zhang
Creator and core maintainer of Yuxi; Ph.D. student in Software Engineering
Jiangnan University Ph.D. student; researcher in NLP, knowledge graphs, entity-relation extraction and LLMs. His GitHub profile says he initiated and built Yuxi's overall architecture and core functionality from zero to one.
Founding Story
This is a personal research and open-source development portfolio rather than a disclosed company. Zhang's stated work centers on NLP, knowledge graphs and LLMs; Yuxi extends that work into a self-hosted knowledge-agent platform combining retrieval, graphs, multi-agent execution and team governance.
Business Model
Revenue Model
Yuxi is distributed as MIT-licensed open-source software for self-hosting; the reviewed official materials describe deployment and contribution rather than a paid hosted service. Users supply their own model API access and infrastructure.
Pricing Tiers
MIT license; run via Docker Compose. Infrastructure, storage and external model/API costs remain with the deployer.
Target Markets
- Enterprise and technical teams needing private deployment
- Individual developers and open-source contributors
- Universities and research teams
- Organizations working with internal documents, knowledge graphs and agent workflows
- Private enterprise knowledge bases and internal RAG
- Knowledge-graph question answering and research
- Multi-agent business workflows and long-running task execution
- Teams requiring control of data, models and permissions
- Document parsing, OCR, retrieval evaluation and file-producing agents
- Researchers working on relation classification and entity-relation extraction