Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Hi there, I’m Adrian Xu. I’m a Senior Software Engineer with over 10 years of experience building AI-powered imaging and production pipelines for packaging and retail. I specialize in AI image generation, prompt engineering, and workflow design using tools like Google Studio, PyTorch, TensorFlow, and Stable Diffusion to produce print-ready artwork while respecting bleed, trim, safe zones, CMYK ICC profiles, and other production constraints. I enjoy turning user notes into production-ready assets and collaborating with design and printing teams to scale robust, maintainable pipelines. On the backend, I build automated QA, asset-management, and data pipelines, and I design cloud-enabled workflows (Docker, GCP, AWS) to support image ingestion, metadata handling, and export formats (PDF/X-1a). I thrive in cross-functional environments, documenting processes, and training teams to ship high-quality outputs faster.…Hi there, I’m Adrian Xu. I’m a Senior Software Engineer with over 10 years of experience building AI-powered imaging and production pipelines for packaging and retail. I specialize in AI image generation, prompt engineering, and workflow design using tools like Google Studio, PyTorch, TensorFlow, and Stable Diffusion to produce print-ready artwork while respecting bleed, trim, safe zones, CMYK ICC profiles, and other production constraints. I enjoy turning user notes into production-ready assets and collaborating with design and printing teams to scale robust, maintainable pipelines. On the backend, I build automated QA, asset-management, and data pipelines, and I design cloud-enabled workflows (Docker, GCP, AWS) to support image ingestion, metadata handling, and export formats (PDF/X-1a). I thrive in cross-functional environments, documenting processes, and training teams to ship high-quality outputs faster.

Adrian Xu

AI Engineer, Data Scientist, Full Stack Developer, +2





Hi there, I’m Adrian Xu. I’m a Senior Software Engineer with over 10 years of experience building AI-powered imaging and production pipelines for packaging and retail. I specialize in AI image generation, prompt engineering, and workflow design using tools like Google Studio, PyTorch, TensorFlow, and Stable Diffusion to produce print-ready artwork while respecting bleed, trim, safe zones, CMYK ICC profiles, and other production constraints. I enjoy turning user notes into production-ready assets and collaborating with design and printing teams to scale robust, maintainable pipelines. On the backend, I build automated QA, asset-management, and data pipelines, and I design cloud-enabled workflows (Docker, GCP, AWS) to support image ingestion, metadata handling, and export formats (PDF/X-1a). I thrive in cross-functional environments, documenting processes, and training teams to ship high-quality outputs faster.…Hi there, I’m Adrian Xu. I’m a Senior Software Engineer with over 10 years of experience building AI-powered imaging and production pipelines for packaging and retail. I specialize in AI image generation, prompt engineering, and workflow design using tools like Google Studio, PyTorch, TensorFlow, and Stable Diffusion to produce print-ready artwork while respecting bleed, trim, safe zones, CMYK ICC profiles, and other production constraints. I enjoy turning user notes into production-ready assets and collaborating with design and printing teams to scale robust, maintainable pipelines. On the backend, I build automated QA, asset-management, and data pipelines, and I design cloud-enabled workflows (Docker, GCP, AWS) to support image ingestion, metadata handling, and export formats (PDF/X-1a). I thrive in cross-functional environments, documenting processes, and training teams to ship high-quality outputs faster.

Available to hire

Hi there, I’m Adrian Xu. I’m a Senior Software Engineer with over 10 years of experience building AI-powered imaging and production pipelines for packaging and retail. I specialize in AI image generation, prompt engineering, and workflow design using tools like Google Studio, PyTorch, TensorFlow, and Stable Diffusion to produce print-ready artwork while respecting bleed, trim, safe zones, CMYK ICC profiles, and other production constraints. I enjoy turning user notes into production-ready assets and collaborating with design and printing teams to scale robust, maintainable pipelines.

On the backend, I build automated QA, asset-management, and data pipelines, and I design cloud-enabled workflows (Docker, GCP, AWS) to support image ingestion, metadata handling, and export formats (PDF/X-1a). I thrive in cross-functional environments, documenting processes, and training teams to ship high-quality outputs faster.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Intermediate

Work Experience

Senior Software Engineer at BrightOps

June 1, 2023 - Present

Designed and implemented end-to-end AI image generation workflows to produce print-ready label art. Enforced bleed, trim, and safe-zone constraints for multiple templates using Google AI Studio, PyTorch, TensorFlow, and Stable Diffusion. Implemented templated prompts that translate user notes into consistent prompts, boosting first-pass print-readiness. Converted RGB renders to CMYK with embedded ICC profiles and produced PDF/X-1a exports at 300 DPI for various print formats. Scaled batch generation to 10,000 background renders per day for peak campaigns. Built monitoring, validation, and QA checks, and trained 8 engineers and 6 designers on best practices and production constraints. Documented workflows and asset metadata schemas; automated export pipelines for 12 label formats.

Machine Learning Engineer at Hexagon Manufacturing Intelligence

January 1, 2019 - May 1, 2023

Built end-to-end generative pipelines for producing texture and background artwork using PyTorch and diffusion models. Achieved 78% art acceptance in internal proofs. Designed and expanded prompt-engineering libraries, data-sets curation, and synthetic augmentation to grow training data by hundreds of thousands of images. Implemented production-grade evaluation criteria, model checkpoints, and color-calibration steps to align outputs with CMYK workflows. Enabled weekly model rollouts, scaled to 1 0 0 0 renders/day, and reduced time-to-market for design iterations. Collaborated with design and print teams to codify label-template standards and created documentation for repeatable production pipelines.

Software Engineer at Progress Software

February 1, 2018 - December 1, 2018

Backend services to support image pre-processing and metadata workflows using Node.js and Express. Improved throughput for image ingestion by 53% and implemented RPIs for asset metadata and versioning. Built automated export pipelines (including PDF-related formats) and REST APIs to support downstream printing partners.

Machine Learning Engineer at NVIDIA Research

November 1, 2017 - October 1, 2018

Explored generative architectures and fine-tuning strategies for image synthesis using TensorFlow and PyTorch. Improved sample coherence by 29% in A/B tests. Prototyped inference pipelines in Docker and Kubernetes and documented prompt-effect relationships across 120 configurations to guide production deployments.