FAQ

I mainly work around backend engineering, platform development, AI Engineering, indie tools, and long-term technical writing. In short, I use backend engineering, AI Engineering, and automation tools to turn complex problems into runnable, reviewable, long-lived systems.

Yes. Java backend and platform engineering are still my technical foundation. I am simply extending microservices, data pipelines, DevOps, reliability, and engineering experience into AI applications, agents, real-time voice modules, and automation tools.

Because real AI delivery quickly becomes a backend and platform problem: service abstraction, authentication, streaming responses, context management, tool permissions, tracing, evaluation regression, cost governance, and deployment stability. Backend capability becomes the foundation for turning AI into systems.

Calling a model API completes a request. AI Engineering asks whether the full system can run reliably. It includes input-output contracts, context strategy, tool calling, permission boundaries, fallback paths, observability, evaluation, cost, latency, and continuous iteration.

They are now separate. The AI Voice Module is an edge-cloud AI interaction system focused on STT, LLM, TTS, agents, interruption, real-time pipelines, and admin configuration. AscendLab is the indie tools and AI Workflow lab for browser tools, small products, and public workflow experiments.

AscendLab is my indie tools and AI Workflow lab. It validates small public tools, AI-assisted flows, and indie product ideas. It is not presented as a mature SaaS. The focus is building useful, accessible, maintainable, and iterated public works.

AI Infra Handbook is my long-term handbook for AI infrastructure knowledge, covering model serving, RAG, vector databases, GPU resources, Kubernetes, observability, Agent / MCP, and cost governance. It is learning and practice notes, not unverified production best-practice claims.

stop-slop-zh is a Chinese AI writing quality governance project. It is not anti-AI; it is against vague, mechanical, over-templated, and fake-depth writing. It can support blog optimization, prompt engineering, Claude / ChatGPT Skills, Chinese content review, and technical writing checks.

It covers backend engineering, microservices, platform development, AI Engineering, LLM applications, Agent / MCP, real-time voice pipelines, indie tools, automation, SEO / GEO, project reviews, and long-term personal growth. Historical posts remain available, while the home and project pages emphasize the current main tracks.

AI will significantly change development workflows, but it does not automatically replace engineering judgment. The more important abilities become problem definition, system boundary design, result verification, quality governance, tool integration into real workflows, and ownership of final delivery.

The projects are not meant to look numerous. They map to different capability layers: wzk.icu is personal infrastructure, AscendLab is an indie tool lab, the AI Voice Module is AI engineering practice, AI Infra Handbook is a knowledge system, and stop-slop-zh plus the code generator are tool-building experiments.

It is not a one-off portfolio. It is a public archive that keeps accumulating. Articles record thinking, projects prove capability, FAQ explains positioning, and llms.txt plus structured data help search engines and AI systems understand the site. Every part supports long-term credibility.

It is for backend engineers, platform engineers, developers moving into AI Engineering, people interested in Agent / MCP / real-time voice systems, indie tool builders, personal technical brand builders, and anyone combining long-term writing with project construction.

To understand me quickly, start with the home and about pages. For AI Engineering, read the LLM engineering article and the AI Voice Module. For backend engineering, go to the Java / microservices series. For projects, start with the project map. For Chinese writing quality, begin with stop-slop-zh.