Before the detail, here's the challenge you'd help us solve.
We build the embodied intelligence that moves real vehicles safely, and the ecosystem a billion machines will run on in the future.
Very few people in AI can say this.
Every role here, whatever the team, plugs into that.
Here’s what this particular role covers.
Gaia is trained on large-scale driving video to predict future frames from past context, functioning as a simulator that generates synthetic scenarios and operates in closed loop with the driving model itself.
The team partners closely with Research, Applications, Core Simulation Engineering, and Cloud/Infrastructure to turn model improvements into measurable downstream impact on the driving stack.
🧠 Your day-to-dayOwn and lead parts of large-scale training for Gaia (language- or video-style training at scale), from experimentation through productionised training runs.
Contribute to and influence model architecture decisions — not just applying existing models.
Partner across sub-teams: Research (model development), Applications (adapting models to use cases), Core Simulation Engineering, and Cloud/Infrastructure.
Translate product and evaluation needs into model improvements (e.
better long-horizon prediction, scenario generation quality, failure-case coverage).
Mentor and set technical direction, raising the engineering and research quality bar across the team.
🧩 What you’ll be working onLeading and executing Gaia's post-training and closed-loop pipeline — fine-tuning and aligning the world model through post-training experimentation and targeted data curation.
Pushing Gaia's autoregressive generation towards longer, more stable rollouts, and making the model deployment-ready (inference time and reliability included).
Contributing to broader model architecture and training-strategy decisions where they inters.