doAZ
Website:
doaz.ai
Job details:
Senior Computer Vision Engineer — Industrial AI Platform Doaz Corp. | Full-time | Remote (India)
About Doaz
We build vertical AI for engineering-intensive industries — construction, shipbuilding, heavy industries, and geotechnical engineering. Our systems are trusted by POSCO E&C, Samsung Heavy Industries, Doosan Enerbility, and KT Estate. We operate at 95%+ accuracy on problems generic AI cannot solve.
Founded: 2023 | Team: 20+ | Mission: Industrial Knowledge to Actionable AI
Who This Role Is For
We hire only engineers with a proven, verifiable track record at the highest level. This means one or more of the following:
- Peer-reviewed publications in top-tier venues (CVPR, ICCV, ECCV, NeurIPS, AAAI)
- CV systems deployed at scale in production environments processing 10,000+ documents or images per day
- Significant open-source contributions in the CV/Deep Learning space with demonstrable real-world adoption
- 5+ years of hands-on engineering leadership in a research-grade or industrial AI environment
The Role
You will own the visual intelligence layer of our industrial AI platform systems that read engineering drawings, extract structured data, and automate compliance verification at production scale. Our models process thousands of real engineering documents daily. A 1% accuracy gain translates directly to millions in saved engineering hours for our clients.
What You'll Build
Drawing Intelligence: object detection and segmentation for engineering drawings, symbol recognition across KS, ISO, and ASME standards, automated cross-reference extraction
Document Understanding: layout analysis, table extraction from BOMs and equipment lists, multi-format processing across PDF, scanned images, and DXF/DWG
Compliance and Verification: regulatory visual verification, design vs. as-built comparison, defect detection pipelines
Domains: construction, shipbuilding, heavy industries, geotechnical engineering
Requirements
Mandatory
- BS/MS/PhD in Computer Science, AI, or related field — MS or PhD strongly preferred
- 5+ years in Computer Vision and Deep Learning with verifiable production deployment at scale
- PyTorch, object detection architectures (YOLO, DETR, Mask R-CNN)
- Strong Python, Git, Docker, Linux
- Demonstrated ability to lead projects end-to-end and mentor junior engineers
- GitHub portfolio, publications, or deployed systems that can be reviewed — applications without a verifiable portfolio will not be considered
Preferred
- Document AI: OCR, layout analysis, table extraction (LayoutLM, Donut, PaddleOCR)
- CAD or engineering drawing domain knowledge
- Industrial or regulated environment experience
- Vision-Language Model research or applied work
- MLOps experience, peer-reviewed publications, or significant open-source contributions
Tech Stack
PyTorch · HuggingFace · Detectron2 · ONNX · TensorRT · OpenCV · PaddleOCR · LayoutLMv3 · Donut · AWS · Docker · Kubernetes · PostgreSQL · Pinecone/Milvus
Why Doaz
Real production impact — not demos, not prototypes
Deep technical work — 95%+ accuracy is our baseline, not our ceiling
Learn from domain experts with decades of industrial engineering experience
Global exposure — Korea-based clients, active expansion into Southeast Asia and the Middle East
Early-stage equity and competitive compensation benchmarked to India's top-tier AI market
Compensation Based on Experience
Target range: $2,500–$3,000 USD/month, commensurate with experience.
Work Arrangement
Fully remote.
Interview Process
- Application Review — Resume, portfolio, and brief statement of interest
- Technical Assessment — Take-home challenge
- Technical Interview — Deep-dive discussion with the engineering team
To Apply
Send your resume, GitHub or publication links, and a short note on why industrial AI interests you to: namal.zafar@doaz.ai
Doaz Corp. — Building AI that understands how industries actually work.
Click on Apply to know more.