Senior AI Engineer / Computer Vision Expert for Engineering Drawing Revision Detection (POC)

We are looking for an experienced AI Engineer with expertise in Computer Vision, OCR, Document AI, and Large Vision Models (VLMs) to help us build a Proof of Concept (POC) for an AI-assisted engineering document comparison platform. The objective is to compare two revisions of technical engineering drawings and automatically identify, classify, and summarize changes to assist engineering review workflows. This is an AI-focused project involving image processing, document intelligence, and structured information extraction rather than traditional web development. The engagement will begin as a short-term POC and may extend into a larger product development opportunity based on the outcome. Project Objectives The consultant will help design and implement a solution capable of: Comparing two versions of 2D engineering drawings (PDF/Image) Detecting visual differences between revisions Extracting engineering annotations using OCR Identifying changes in dimensions, notes, annotations, symbols, and metadata Classifying detected changes into meaningful categories Producing a structured change report Designing an extensible AI architecture suitable for future enhancements Scope of Work The consultant will be expected to: Discovery & Solution Design Review sample engineering drawings Recommend an overall AI architecture Identify technical challenges and mitigation strategies Define the POC implementation approach AI & Computer Vision Image preprocessing Drawing alignment Difference detection OCR integration Region detection Engineering entity extraction AI-assisted change classification Deliverables Working POC Source code Architecture documentation Technical documentation Deployment instructions Knowledge transfer session Preferred Technical Skills Strong experience in several of the following: Programming Python Computer Vision OpenCV scikit-image Image registration Feature matching Object detection OCR / Document AI PaddleOCR Tesseract EasyOCR PDF processing AI / LLM / Vision Models Experience with one or more: Qwen-VL LLaVA Florence GPT Vision Gemini Vision Other Vision Language Models Frameworks FastAPI Streamlit Gradio Document Processing PyMuPDF pdf2image Pillow Nice to Have Experience with: Engineering drawings CAD documentation Manufacturing documentation Technical documentation workflows Industrial AI Document Intelligence Layout analysis Diagram understanding Prior experience with engineering or manufacturing documentation is a plus but not mandatory. Expected Deliverables AI solution architecture Working prototype Image comparison engine OCR pipeline Change detection engine Classification logic Review interface (basic) Exportable structured report Source code with documentation What We Are Looking For The ideal consultant should: Have hands-on experience building AI products (not only research) Be comfortable working with document AI and computer vision Understand practical AI limitations and propose realistic solutions Be able to recommend open-source technologies where appropriate Communicate technical concepts clearly Engagement Initial engagement: POC (4–6 weeks) Potential extension into a long-term product development engagement based on successful delivery.