FOUNDER

Hequbing · He Fangsheng · Dosen

I combine business experience with hands-on software work. Hequbing (贺去病) and He Fangsheng (贺方升) refer to the same person; Dosen is another name I use.

Background

My career spans around 20 years, beginning with search technology, followed by entrepreneurship, logistics-related business and international medical-supply market development. I now focus on AI business consulting and product development.

That experience informs two areas of work: helping owners choose practical AI opportunities, and building tools from problems encountered during implementation. My approach is described in embedded AI implementation.

Projects and practice

These notes describe projects, historical versions, methods and controlled tests. Their evidence and limits are stated in each article.

CAD AI auditing: traceable window and door inventories from DWG and DXF

An object-level CAD workflow records counts, locations and classification evidence, with clear limits on geometry recovery and validation.

Task Passport: handing a task between Claude, Codex and DSH

Task Passport preserves goals, verified facts, artifact references and next steps, while version checks reject stale writes.

OpenCodex: a local multi-terminal workspace for AI coding

A project-folder workspace with split terminals, process lifecycle management, user-configured model endpoints and no embedded API keys.

Building U-King: configuration, processes and real-machine failures

Lessons from configuration write-back, same-name process termination and actual installation-state detection in a multi-tool AI installer.

Hequbing Business Consulting: diagnosis, pilots and delivery

Three stages of AI business consulting, with CNY pricing, agreed scope, acceptance criteria and attribution of incremental results.

Companion FDE: implementing AI with a business team

A field-oriented approach connecting business judgment, implementation and reviews, leaving usable workflows and knowledge with the client.

AnythingChina: connecting logistics experience with AI-assisted sourcing

A project that combines AI research with authorized local contact and follow-up. Its proposed network size is a goal, not established coverage.

U-Hermes portable AI environment: versions, setup and operating requirements

The historical U-Hermes design packaged Hermes Agent and its runtime for portability. Local software, model inference and dependency licensing remain separate questions.

How to evaluate GEO: delivery, discoverability and business results

Evaluate generative engine optimization in three layers: agreed work, observed mentions and citations, and attributable business outcomes. A screenshot is not a performance guarantee.

Public work and identity

Public repositories are available on GitHub: dongsheng123132. The project articles identify the relevant source materials and distinguish project goals from implemented capabilities.

Service scope and pricing cover business consulting and technical delivery separately. For Chinese background information and the broader article archive, use the Chinese language link.

Contact

Email hefangsheng@gmail.com or contact me on WeChat: hecare888. Include your business, the problem you want to solve and any delivery constraints.