Responsibilities may include the following and other duties may be assigned:
- Develop automated PDF verification systems - Text and Images
- Implement hybrid matching algorithms (template + cosine similarity + text context)
- Multi-threaded processing pipelines (thread pooling)
- Extract embedded images from Excel, leverage Adobe source materials
- Design AI agentic workflows for automated document processing
- Build multi-agent systems (Coordinator, Vision Agent, Regulatory Agent, QA Agent)
- Integrate LLMs (GPT-4, Claude, Gemini) for intelligent analysis
- Implement human-in-the-loop validation patterns
- FDA/ISO regulatory compliance documentation.
Required knowledge and experience
- Python expert: NumPy, OpenCV, PyMuPDF, scikit-learn, threading
- Computer Vision: Template matching, cosine similarity, SSIM, multi-scale detection
- 12+ years total experience, 4+ years in CV/image processing
- Medical device regulatory standards (FDA 21 CFR, ISO 13485)
Nice To Have Skill
- LLM Frameworks: LangChain, LlamaIndex, AutoGen, CrewAI
- Prompt Engineering: Design effective prompts for CV tasks
- Tool Use Patterns: Function calling, tool schemas
- Multi-Agent Systems: Agent teams with specialized roles
- RAG: Vector databases (ChromaDB, Pinecone) for context
- HITL: Approval gates, escalation, feedback loops
