该职位来源于猎聘 Objective Design, develop, and deploy the CMD Design Review AI Tool (leveraging LLMs, multi-modal document parsing, and RAG architectures) toautomate mechanical design rule checks, review 2D/3D design packages against CMD standards, and train SI partners on tool adoption. Roles &
Responsibilities (R&R)
- AI Tool Development &
- Architecture: Lead end-to-end development of the automated CMD Design Review AI Tool, including promptworkflows, structured schema extraction, and domain-specific knowledge retrieval.
Multi-Modal Document Parsing: Build and optimize ingestion pipelines for technical drawings, PDFs, mechanical spec sheets, and complexBOM/part tables.
Integration &
- Deployment: Containerize, deploy, and maintain the application in production, implementing logging, observability, andevaluation benchmarks.
SI Training &
- Tool Rollout: Author technical documentation and user manuals; train SI engineering teams to adopt the AI tool in their dailyreview workflows, collect feedback, and iteratively improve accuracy.
Resource
Qualifications &
Software Foundations: 3 years of professional software engineering experience with strong proficiency in Python (or Go / TypeScript /C ). Solid command of Git, Docker containerization, logging & monitoring, and automated testing frameworks.
LLM Application Development: Solid hands-on experience developing LLM-driven systems: o Advanced prompt design and context window management. o Reliable structured outputs via JSON Schema. o Tool use / Function calling (including MCP — Model Context Protocol). o Robust RAG (Retrieval-Augmented Generation) architectures.
Production Delivery Track Record: Proven experience building and shipping at least one production-grade AI application, autonomousAgent, or enterprise knowledge base (capable of demonstrating system architecture, Evals, and real-world failure mitigation).
Unstructured Data &
- Document Extraction: Practical experience with document parsing, OCR pipelines, complex table extraction, andhandling irregular technical drawings/layouts.
Search &
- Database Engineering: Solid background in relational databases alongside practical expertise in implementing and tuning full-textsearch, vector retrieval, hybrid search, and reranking pipelines.
Communication: Fluent in spoken and written English and Mandarin. ________________________________________
Key Deliverables
Regularly updated and released CMD documentation and engineering guideline revisions.
Verified 3D CAD models, 2D drawings, and DFM packages aligned with latest HWTE requirements.
Maintained and up-to-date CMD Global Component List &
Successful integration and deployment of the CMD database into designated SI design environments.
Deployed and verified production release of the CMD Design Review AI Tool.
User manuals, technical documentation, and completed training sessions for SI engineering teams.
