Foundation Model Engineer

About MarqVision

Protecting and building a future shaped by original ideas, innovations, and creativity. From counterfeits and pirated content to fraudulent sites and unauthorized sellers, MarqVision provides an end-to-end brand integrity platform across e-commerce, social media and emerging AI platforms. Our AI-powered managed service platform for brand control and intelligence enables rapid, scalable enforcement so that brands can move faster, protect revenue, and preserve brand equity across every digital touchpoint. In a time when IP infringements pose an increasingly serious challenge, we believe that democratizing legal expertise and access to the latest in generative AI technology is critical for continuing to support the inventors, developers, and artists of the world who work tirelessly to deliver safe, high-quality products. Founded by Harvard Law graduates in 2020, MarqVision recently raised Series B, bringing total funding to $89M, and is backed by Y Combinator, Peak XV (formerly Sequoia Capital India), HSG (formerly Sequoia Capital China), Salesforce Ventures, Altos Ventures, DST Global Partners, and Softbank. Recognized with an LVMH Innovation Award and the Inc 5000 Fastest Growing Companies award in 2025, MarqVision continues to push boundaries in brand control, ensuring that innovation, creativity, and trust remain safeguarded in a rapidly evolving digital world.

Why We're Building This

General-purpose LLMs are powerful but do not consistently understand domain data. MarqVision is solving this problem by building its own foundation model that understands online product and IP information together. We believe that one model serving as a common foundation for multiple pipelines such as detection, classification, and enforcement, and as that model improves, the entire product gets better. We believe the true competitiveness of AI products comes not from features, but from the data understanding shared across those features.

What You'll Achieve

• Design and train a domain-specific foundation model that understands online product images, text, and IP information together.
• Lead the entire process from training data pipeline to model evaluation, connecting model performance to actual product metrics.
• Deploy the trained model to the actual service and integrate it with various AI pipelines to ensure the entire system operates consistently.

Skills You'll Need to Bring

• 2+ years of development experience with LLMs or multimodal models.
• Experience operating end-to-end from model training to serving, and continuously improving based on real service feedback.
• Experience in integrated design within an environment where multiple AI pipelines are linked.

Nice to Have

• Experience with domain-specific VLM pre-training and fine-tuning.
• Experience with distributed training and serving optimization based on GPU clusters.
• Ability to collaborate in English for work.

Interview Process

Please submit all documents as a PDF file

The interview process is as follows:

• Application Review
• Screening Interview
• Hiring Manager Interview
• Team Interview
• Culture Fit Interview
• Case Study (optional)
• Final Interview

The interview process is subject to change.

This is a full-time position with a 6-month probationary (trial) period

Please ensure that all submitted documents are accurate. Any false information may result in the withdrawal of your acceptance or offer of employment.

Equal Opportunity

At MarqVision, we believe real growth comes with attracting the brightest minds, motivating them to succeed in their position, and ensuring plenty of fun along the way. If you think this position is for you, apply now!

MarqVision is an Equal Opportunity Employer and encourages applicants of all backgrounds to apply. We are committed to building a diverse and inclusive workforce that reflects the customers and communities we serve. By applying for this role, you acknowledge MarqVision’s Global Recruiting Privacy Policy. Your privacy is important to us. Please review our Candidate Privacy Notice .