Senior AI Research Scientist (PhD)

Generative Models · Full-time, permanent · Madrid, Spain"Wanted: a researcher for an unsolved problem. Few established methods, no off-the-shelf benchmarks, an industry watching. Papers welcome. Production guaranteed."Meet the team!We are Modelia.We are building the platform that will transform how the fashion industry creates, manages, and scales visual content. For decades, brands have relied on photoshoots as the foundation of their visual production. They are expensive, slow, and deeply embedded in every P&L.We believe that model is coming to an end.Our engine turns a single garment photo (a flatlay, a mannequin shot, an existing model image) into studio-quality on-model images and video. No photoshoot. Minutes instead of weeks. We already work with brands across 26 countries and have generated on-model imagery for more than 1,000,000 SKUs.We are the official AI-imagery partner of Mercedes-Benz Fashion Week Madrid, we are part of IED Milano's fashion-design curriculum, and we are backed by top US and EU investors.We are a focused team, moving fast, building something that doesn't have a clear precedent.The RoleThis is the research seat at the core of the company.You will own the research direction of our generative engine: the image and video models behind controllable generation, editing and personalization.You will decide what to build and why. You will turn open research questions into methods that work. And then you will see them run in production, for real brands, at scale, every day.The problems are hard and specific. How do you keep a garment's fabric, fit and print exactly right? How do you keep a person's identity stable across poses? How do you keep a video consistent frame after frame? There are no established answers. That is the work.Your research will not end in a drawer. It will ship.What you'll doLead the full research cycle for image and video generation: hypothesis, experimental design, implementation and rigorous evaluation.Advance the problems behind on-model generation: garment-conditioned synthesis and virtual try-on, identity and pose preservation, texture and garment fidelity, and temporal consistency in video.Improve the controllability, fidelity and inference speed of the pipeline.Own how we measure quality. Define the metrics and evaluation protocols, combining quantitative scoring with structured human review.Work hand in hand with engineering to take research into production.Publish, present Modelia's research at conferences, and pursue patents where relevant.Mentor junior researchers, including our Industrial PhD candidate.What we're looking forA PhD in Machine Learning, Computer Vision, Applied Mathematics or a closely related field.A strong publication record in image or video generation (CVPR, ICCV, ECCV, NeurIPS, ICLR, SIGGRAPH or equivalent).Rigour. You know how to pick the right metric, run a controlled experiment and read results critically, including the ones you didn't want.Strong engineering skills and experience running large-scale ML experiments and data pipelines.Clear communication in English.You'll have an edge if you bringDepth in video and temporal generation, subject- or garment-driven personalization, image editing and inpainting, or efficient serving of generative models.A track record of turning research into shipped products.International research stays at strong academic or industry labs.Research leadership: PI or work-package lead, secured funding, supervised students, program-committee or reviewing roles.Open-source contributions, granted patents or released datasets the community actually uses.A public GitHub, and a genuine curiosity for fashion and visual content.What We OfferYou will:🧠 Own the research behind a product that is defining a new category. 🚀 See your work in production, used by real brands, not just in a paper. 👊 Work directly with the founders and the engineers, without layers or politics. 📄 Get budget to publish and attend the conferences that matter. 🎓 Mentor and co-supervise our Industrial PhD researcher. 💰 Get a competitive salary and a permanent, full-time contract. 📍 Be based at our office in Madrid, Spain.Good to knowTo be eligible, you must hold a PhD. Candidates of any nationality are welcome; the role itself must be based in Spain.Ready to Help Us Push the Limits of AI in Fashion?Then follow these steps:🤗 Follow Modelia on LinkedIn.🧪 Try our Studio. Generate a few images and see where the models still fall short.📝 Apply via LinkedIn with your CV and your Google Scholar profile (or 3–5 representative papers).✉️ We'll reply within a week with next steps.Applications close on 15 October 2026.Can't wait to meet you 👋!