Two paths times two technology stacks
The path decides how value is created; the technology stack decides what it is created with.
Delivery path (FDE engineer) — make systems work on customer sites
- AI software direction: AI application delivery, agents and RAG, data and system integration
- Embodied-AI direction: robot workstations, edge devices and visual inspection — on-site deployment, calibration and acceptance
R&D path — build products and technology in-house
- AI software engineer: models and algorithms, agent frameworks, toolchain and platform
- Embodied-AI engineer, split into two tracks
- Build track: structure, drive, control and embedded systems — building the robot
- Application track: SDK and ROS 2, sensor calibration, upper-layer algorithms, scenario integration
Four-stage pipeline
| Stage | What happens |
|---|---|
| 1 Attract | Three channels (events, base, universities); classes split by path |
| 2 Cultivate | Common foundation, path-specific modules, project practice |
| 3 Sign | Contracting and talent pool, tagged by direction |
| 4 Dispatch | Secondment to companies, projects or certified positions |
Four programmes
- Competition camps: 4–8 weeks of intensive training and mock contests; strong teams go straight into Chuangsheng Works
- University joint programmes: joint courses, internship base, credit recognition
- Enterprise training: customised to job competency models across four directions
- Certification and course licensing: a four-direction certification system, licensable to institutions
Course structure
Common foundation (Python / ROS 2 / data and model basics) + path modules (delivery: methods, deployment and acceptance, hardware interfaces; R&D: algorithms, platform, application development) + project practice.
Where the classes are
Intelligent Manufacturing Entrepreneurship Park, Hongqi District, Xinxiang, Henan, China.