FOCUS · 01
AI agents & knowledge systems
Design agent, retrieval and tool integrations for document Q&A, information extraction and business workflows.
- Document Q&A and retrieval
- Agent tools and workflows
- Structured extraction and human review
● Open to project enquiries
AI AGENTS / KNOWLEDGE / SOFTWARE
AI agents and
business software.
I’m Yong Zhang. I build AI agents and knowledge systems for document Q&A, information extraction and business workflows, and help with Java backend development, system modernization and deployment.
EXPERTISE & TECHNOLOGYAI Agent·Knowledge / RAG·Java applications·Deployment & operations
01 / SERVICES
From AI agents to business systems
Discuss the actual need first, then shape an approach across design, development and maintenance.
DEVELOP & DELIVER
02
Web application development
Build business applications, admin systems and APIs, from data design through implementation.
- Administration systems and APIs
- Database and third-party integrations
- Web and mini-program backends
03
Modernizing existing systems
Maintain Java applications, improve interfaces and data models, and help evolve monoliths into services.
- Java maintenance and iteration
- Data models and interface review
- Incremental service decomposition
04
Deployment & operations
Support Linux deployments, containers, monitoring, troubleshooting and ongoing maintenance.
- Linux and container deployment
- Monitoring and troubleshooting
- Maintenance and delivery support
02 / SELECTED EXPERIENCE
The work, and my contribution.
Anonymized examples describing the context, modules and my role, without identifying clients or companies.
CASE 01 · AI AGENTS / KNOWLEDGE
Engineering document AI agent
ContextEngineering projects involve many document formats. Classification, information capture and retrieval can require substantial manual work.
Key modulesDocument classification, field extraction, file review and project Q&A (designed and developed by scenario)
My contributionContributed to the design and development of an agent approach using document parsing, retrieval and tool calls, with attention to structured results, human review and business-system integration.
TechnologiesJava · Spring Boot · AgentScope · RAG (local work and design)
Illustrative workflow
Project documents → parsing → classification, extraction or retrieval → human review → Q&A or business systems
Developed in stages; the diagram illustrates the intended workflow, not a claim that every step is in production.Example scenarioFor example, a project team needs to organize a batch of files: identify document categories, extract fields for review, and use relevant material to answer project questions after human confirmation. The actual workflow depends on the data and system.
Contribution and scopeContributed to agent orchestration, structured output and integration design. Production identity, retrieval and system integration require project-specific validation.
Have a similar project? Get in touch →CASE 02 · AI AGENTS / KNOWLEDGE
Knowledge-base Q&A and customer support
ContextPolicies, course materials and common questions are distributed across documents, making repeated enquiries harder to manage.
Key modulesDocument knowledge base, retrieval Q&A, conversational workflows, fallback
My contributionExplored Dify knowledge bases and dialogue flows, including document chunking, retrieval, business API integration and handling missing information or human follow-up.
TechnologiesDify · Knowledge base · Chatflow · REST API · Docker
Illustrative workflow
Knowledge documents → chunking → retrieval Q&A → human follow-up when needed
Exploratory solution; not presented as a production deployment.Have a similar project? Get in touch →CASE 03 · Engineering practice
Online learning and examination platform
ContextLearning and examination services need to support multiple client applications and organize question banks and exams.
Key modulesUsers and access, question banks, exam assembly, learning records
My contributionContributed to multi-client APIs and related features, organizing data and workflows for questions, exams and answers.
TechnologiesJava · Spring Boot · MySQL · Redis · RabbitMQ
CASE 04 · Engineering practice
Charging device and operations system
ContextCharging services need to connect devices, process messages and support operational and agent management.
Key modulesDevice communication, device data, agent management, payments and settlement
My contributionHelped maintain and improve device reporting, statistics, payment and agent functions, including the handling of duplicate messages.
TechnologiesSpring Boot · Netty · PostgreSQL · Redis · RabbitMQ
CASE 05 · Engineering practice
Repair service and work-order platform
ContextA repair service needs to connect customers, technicians and an administration team through an online workflow.
Key modulesService requests, technician management, work orders, administration
My contributionContributed to workflow design and the initial mini-program and management application, including access control, service management and payment integration.
TechnologiesSpring Boot · Redis · MariaDB · RabbitMQ · WeChat Mini Program
CASE 06 · Engineering practice
Education scheduling and course management
ContextScheduling requires coordination of teacher availability, courses and administrative work while maintaining existing systems.
Key modulesCourse management, teacher availability, scheduling, users and access
My contributionParticipated in requirements analysis, data-model improvements and maintenance, and helped split selected functions from a monolith into services.
TechnologiesSpring Cloud Alibaba · Nacos · MySQL · Redis · RabbitMQ
03 / ABOUT
Direct communication, hands-on work.
I’m Yong Zhang. My background is in Java backend and business application development, with work using Python and Go. I focus on AI agents, document knowledge systems and connecting them to existing workflows, supported by development and operations experience.
Scope and capabilities are discussed for each project; confidential experience is presented anonymously.
Deliverables can include source code, API documentation and deployment handover, as agreed for each project.
View a public GitHub project ↗
Get in touch ↗04 / WORKING TOGETHER
Make the scope clear, step by step.
Understand
Discuss goals, constraints and existing systems.
Define
Agree on a practical scope and delivery priorities.
Build
Develop within the agreed scope and share progress.
Hand over
Prepare the deliverables and discuss ongoing support.
05 / TECHNOLOGIES
Choose tools to fit the problem.
These are technologies I have used across development and systems work.
Java / Python / Go · Spring Boot / Spring Cloud · Vue · MySQL / PostgreSQL / Redis · RabbitMQ · Docker / Linux / Nginx
06 / FAQ
Questions before we start.
Can you work on an existing system?
Yes. We can first discuss the current system, the problems and the scope of support.
Can you help with an AI agent or knowledge base?
We can discuss document Q&A, information extraction and integrations; the scope depends on the project and available data.
How do we get started?
Send the requirements, existing context and desired timeline by email, Telegram or WeChat so we can discuss feasibility.
How are price and timeline determined?
They depend on scope, existing systems, data readiness and delivery requirements and are agreed after discussion.
07 / GET IN TOUCH
Have a problem to solve?
Let’s talk.
Share what you are building, your current situation and your timeline. Reach me by email, Telegram or WeChat.
