Neoteny AI is building the sovereign intelligence layer for code. We are a team of builders from Meta, AWS, and NYU who believe that the future of coding intelligence is not a generalist chatbot in the cloud, but a specialized, repository-aware model that lives inside the enterprise perimeter. We are looking for a founding researcher to lead our core research agenda. You will work at the intersection of representation learning, program synthesis, and efficient inference. You will figure out how to teach models the latent physics of a codebase, the implicit architectural rules, dependency patterns, and style constraints that generalist models miss. Your work will range from designing new data engines to experimenting with novel architectures that break the memory wall of current transformers. We are looking for someone with deep ML expertise (PhD or equivalent) who is code-fluent and comfortable writing kernels, not just training scripts.