Synthetic Data Generation Engineer (Computer Vision & Graphics)

Deepfakes and synthetic data are no longer niche technologies: they're mainstream, and with that comes both immense potential and real risk. At Microblink, we believe in using these powerful tools for good. Would you like to join a team of fraud analysts, machine learning engineers, and computer vision experts working at the frontlines of identity fraud prevention?

We are looking for a Synthetic Data Generation Engineer to help us fight fraud and reduce bias in our AI systems. You’ll play a critical role in developing scalable, realistic, and diverse synthetic datasets for training and evaluating computer vision models. Your work will directly support identity document verification, biometric authentication, and fraud detection systems used globally.

If you're passionate about traditional computer graphics tools like Photoshop and Blender, always following the news in generative AI (e.g., Hugging Face models Hub), and have a hacker mindset for breaking and improving systems, this is the perfect opportunity for you! You’ll have the unique chance to build synthetic data pipelines from scratch, experiment with cutting-edge generation techniques, and contribute to launching ML-powered fraud prevention products that impact millions of users worldwide.

A little bit about us

Microblink is an AI company specializing in computer vision and identity fraud detection. We create unique products and tools designed to make the future of digital identity a secure "no-fake-zone." Our expertise includes running efficient machine learning models directly on-device, enabling a uniquely seamless Identity Verification (IDV) experience. With teams across the US and Europe and main offices in New York and Zagreb, we touch hundreds of millions of users each year, processing over 800 million identity documents across more than 70 countries worldwide. Explore our tech in action here.

Here’s what you’ll do

• Design and implement scalable pipelines for generating synthetic images and videos for training computer vision models.
• Support and develop our in house data generation platform
• Leverage both traditional CG tools (e.g., Photoshop) and AI-based generative models (e.g., diffusion models).
• Collaborate closely with fraud experts and ML engineers to understand real-world attack vectors and simulate them synthetically
• Showcase your work in public benchmarks and through demos with our clients. Your communication skills will be key!
• Experiment with new techniques for generating diverse synthetic identities that can improve model robustness and fairness.
• Contribute to reducing demographic and geographic biases in identity verification systems through smart data augmentation and generation strategies.
• Bring a "hacker mindset" to outthink fraudsters by simulating edge cases, adversarial attacks, and novel fraud patterns.
• Help the team make data-driven, sound decisions regarding new technologies, standards and architecture

You'll be successful in this role if you

• Have practical experience with 2D tools like Photoshop, and/or experience with 3D rendering or CG pipelines (Blender, Unity, Unreal)
• Have programming skills in Python, with knowledge of image processing (e.g., Pillow) and computer vision (e.g., OpenCV) packages.
• Have a creative, experimental mindset with a strong sense of mission to outsmart fraud and build fairer AI systems. Attention for details in generating and validating data is key here.
• Have an essential software development skillset (“it goes without saying” type of skills): git, team collaboration skills and tools, and good English language skills
• Have excellent communication skills
• Are responsible, proactive and driven by organizational and team success
• Take full ownership of your work from idea to delivery, including design, implementation, testing, and maintenance

A plus is if you also

• You are curious about Image processing and Computer Vision and deeply passionate about what’s new in the generative AI space (e.g., diffusion models, multimodal LLMs). We’d love to hear that you spend too much time on Hugging Face models Hub, MidJourney, ComfyUI, or Vertex Model Garden.
• Have experience in front-end development and JS frameworks like Vue.js
• Containerization tools like Docker
• Cloud environment, such as GCP

Here’s what you’ll gain if you join us

• Microblink owes its success to its people. That is why we offer equity participation to all of our full-time employees so that we can grow together
• Unlimited PTO, giving you the flexibility to take time off as needed to maintain a healthy work-life balance.
• Location-specific "Pick & Choose" benefits tailored to meet diverse personal preferences and regional needs.
• Flexible work arrangements, allowing you to adjust your schedule—whether working remotely or at our Zagreb office.
• Opportunities to learn and develop your skills through internal L&D programs, and a supportive working environment.

How can you join us?

If this role intrigues you, do not hesitate to hit the apply button and send us your resume. Get in touch with us, and we promise to get back to you

Microblink LLC is an equal opportunity employer committed to diversity and inclusion in the workplace. This policy applies to all employment practices within our organization, from hiring to the end of your business relationship with us. Accordingly, this job advertisement is equally applicable to all people and Microblink will make a hiring decision based solely on qualifications, merit, and business needs. We prohibit discrimination of any kind based on race, color, sex, religion, sexual orientation, national origin, disability, genetic information, pregnancy, or any other protected characteristic as outlined by federal, state, or local laws, as well as our internal policies and culture.

For detailed information on how we process personal data, please read our Privacy Policy https://microblink.com/privacy-policy/ or reach out to our Data Protection Officer at [email protected].