Software development
Custom web applications, admin backends, mobile apps, and mini programs — from requirements through launch. Clear structure, reproducible deployments, and code that stays easy to iterate on.
View details →Four services covering the stages a product goes through — from zero to launch, and from launch to long-term operation. Engage us for one, or combine them.
Custom web applications, admin backends, mobile apps, and mini programs — from requirements through launch. Clear structure, reproducible deployments, and code that stays easy to iterate on.
View details →Servers, containers, and CI/CD pipelines, plus monitoring, alerting, backups, and security hardening. Systems that stay up, and documentation that lets someone else take over.
View details →Self-hosted LLM deployment, retrieval-augmented knowledge bases, support automation, and workflow automation. Model capability wired into a specific job — not just a demo.
View details →Model selection, RAG and agent architecture, prompt tuning, code review, and day-to-day engineering questions. Pay per question or monthly, and ask as soon as you check out.
AI engineering consultation can be ordered online: ¥6 per question, ¥66 per month. Custom development and managed operations are quoted per project through a one-to-one conversation.
Custom applications: web / mobile / mini programs
Server management, monitoring, alerting, security hardening
An engineering advisory subscription for teams building with AI
Requirements analysis, architecture, development, testing, and deployment, run as verifiable iterations. Ship the part that works first, then keep improving it.
Monitoring from infrastructure up to the application layer, reproducible deployment processes, and operations documentation that is actually readable. Fast diagnosis when something breaks, and a clean handover when someone leaves.
Self-hosted LLM deployment, retrieval over your own documents, and application development. We define the use case and how it will be evaluated before choosing models, data, and architecture.
When you hit questions about model choice, API integration, prompt quality, performance, or cost while building with AI, an experienced engineer answers one to one. Ask a single question, or subscribe monthly and ask whenever.
From the first conversation to delivery, every step has a defined output and a sign-off point.
Understand the business context, goals, and constraints, and pin down scope and acceptance criteria.
A technical approach, a schedule, and a price — with inclusions and exclusions written down.
Runnable results delivered in stages, with regular progress updates and course corrections when needed.
Operations and iteration support after launch, so the system keeps running.
Call or add us on WeChat, tell us the background and the goal, and we will work out the next step together.