AI Researcher — Foundation Models for Floorplan Design

Meetdavis; Paris; Hybride; EUR 90 000 - 130 000; Plein temps

Avantages offerts par ce poste

Competitive salary

Meaningful equity

Paris office

Résumé du poste

Davis is building a foundation model for architectural design that creates compliant, editable floorplans from scratch. You will lead end-to-end research from diffusion models to deployment, collaborating with architects and a small team of engineers.

You will own model architecture, large-scale training on GPU clusters, and benchmarking to surpass state-of-the-art results, shaping the tools used by real estate professionals.

Qualifications

• PhD or Master’s in Maths, Computer Science, ML, or related field, or equivalent experience.
• Strong foundations in ML and a track record of innovative research.
• Diffusion model expertise (discrete or continuous) and guided generation.
• Experience training large-scale DL models on GPUs and distributed systems.
• Proficiency in Python with PyTorch or similar frameworks.
• Ability to read, evaluate, and implement advanced ML research.
• Excellent collaboration and communication skills.

Responsabilités

• Design the core model architecture and a discrete design space for architectural layouts.
• Lead the training of a foundation diffusion model on GPU clusters and manage experiments at scale.
• Benchmark, iterate, and push performance beyond existing methods to deliver deployable architectural designs.

Connaissances

Applied ML Research

Diffusion Models

Large-Scale Training

Python & PyTorch

Distributed Training

Research Literacy

Collaboration & Communication

Formation

PhD or Master’s in Maths, CS, ML

Outils

PyTorch

Description du poste

Davis is building a foundation model for architectural design that creates compliant, editable floorplans from scratch. You will lead end-to-end research from diffusion models to deployment, collaborating with architects and a small team of engineers.

You will own model architecture, large-scale training on GPU clusters, and benchmarking to surpass state-of-the-art results, shaping the tools used by real estate professionals.