AI Research Intern

Overview

Davis AI, an AI-native real estate company that speeds up early-stage development and architectural design, is hiring an AI Research Intern in Paris to push forward its floorplan generation model. The company argues that image-based models produce plans that break building codes under scrutiny, while its model works on structure rather than pixels and already beats the benchmarks it has been tested on. The foundation model uses discrete diffusion on structured representations to generate compliant, editable floorplans and site plans that respect constraints such as zoning rules and space requirements. The intern joins a focused team of three engineers and researchers, within a company of 12, and works daily with architects.

What you will do

The intern takes open questions about the model, reads the literature, forms hypotheses and designs ablations that can falsify them. Work also includes improving the model through multi-GPU training runs, reading loss curves and generated samples and iterating on architecture, loss design, conditioning and sampling. A third strand is dataset creation: building an agentic, computer vision based pipeline that parses raster plans, PDFs and scans into a clean structured representation for training.

Requirements

Candidates should be in a top MSc or PhD programme, or equivalent, in machine learning, computer science, applied mathematics or a related field, with a strong understanding of discrete or continuous diffusion models and guided generation. Strong Python and PyTorch skills are needed, and familiarity with distributed training libraries such as PyTorch Lightning is a plus. The employer also wants the ability to read and implement advanced research and to communicate clearly with engineers, researchers and domain experts.

Preferred extras include a previous internship in diffusion or generative modelling, publications or open-source contributions, interest in architecture, experience with reinforcement learning and optimisation, graph-based or layout generation, document and image understanding, and LLM agent pipelines.

What you get

The role offers a competitive salary and meaningful equity in an early-stage company, with fast progress from research to prototype to production.