Research Lead / Principal Scientist & Manager Post-Training · Alignment · Reinfo

Job Requisition ID #26WD98298Research Lead / Principal Scientist & ManagerPost-Training · Alignment · Reinforcement LearningAutodesk AI Lab: London · San Francisco · Toronto · Remote (US/CA/EU)Open to Remote: Germany, France or ItalyThe OpportunityFoundation models are reshaping how engineers, architects, and designers work —but training foundation models that are reliable, domain-capable systems is still an open research problem.Autodesk touches more of the physical world than almost any other software company. The products we build are used to design skyscrapers, manufacture aircraft, and produce films. AI is now central to how those workflows are evolving — and post-training is the layer that makes the difference between a capable model and one that is dependable and robust in our customers’ high-precision domains.As Research Lead for Post-Training & Alignment, you will own Autodesk's research strategy for transforming foundation models into systems that are reliable, aligned, and genuinely useful in complex, domain-specific workflows. This is a deeply technical leadership role — you will shape research direction, drive key architectural decisions, and remain close to the work.You will lead a growing team of AI scientists while continuing to contribute directly to research: running experiments, developing novel algorithms, and publishing at top-tier venues.This role reports to the Senior Director of AI Research within Autodesk AI Lab.Why This RoleUnique research surface areaAutodesk's domains — architecture, engineering, construction, manufacturing, media & entertainment — provide a distinctive research environment: rich structured data, long-horizon reasoning tasks, and real-world evaluation grounded in professional workflows. Uniquely, decades of investment in physics simulation engines, CAD kernels, and computational design tools give us something most labs don't have: high-fidelity, domain-grounded verifiers that can serve as reward signals for post-training. Rather than relying solely on human preference data, we can ground reinforcement learning in the laws of physics and the constraints of real engineering. These are exactly the kinds of challenges — and assets — that make post-training and alignment research here genuinely distinctive.Research-first, with real impactWe publish at NeurIPS, ICML, ICLR, CVPR, and SIGGRAPH. We collaborate with leading academic and industry labs. And we have a direct line from research advances to product impact at scale. This is not a role where research sits behind a wall from engineering — you will see your work matter.What You Will DoResearch & Technical LeadershipOwn post-training strategy for model development — from RLHF and preference optimization to agentic systems and long-horizon reasoningDevelop novel algorithms that improve model reliability, controllability, and alignmentMake principled architectural decisions about when to address challenges at the pre-training, post-training, or system levelDesign and run experiments that shape model behavior, robustness, and reasoning qualityPartner with infrastructure teams to build scalable, reproducible post-training workflowsContribute to publications, patents, and Autodesk's external research visibilityEvaluation & Model QualityDesign evaluation frameworks for long-horizon reasoning, tool use, agentic behavior, safety, and real-world workflow completionLead rigorous model analysis and interpretability effortsDrive human-in-the-loop evaluation with high annotation quality and sound scientific methodologyEstablish model readiness criteria and provide go/no-go recommendations for releasesCommunicate technical risks, limitations, and trade-offs clearly to leadershipTeam & Organizational LeadershipManage, mentor, and grow a team of AI scientistsSet technical direction and research priorities across post-training and alignment initiativesFoster a research culture grounded in scientific rigor, reproducibility, and fast iterationHelp recruit world-class talent across ML, RL, alignment, and foundation modelsPartner closely with pre-training teams, infrastructure, product organizations, and other stakeholdersTranslate research trade-offs into clear, decision-ready guidance for leadershipWhat We Are Looking ForWe care about research judgment and outcomes, not credential checklists. Strong candidates will typically have:Deep hands-on expertise in reinforcement learning for foundation models, and fluency with post-training methods (RLHF, RLAIF, DPO, PPO, or adjacent approaches)Proven experience leading or mentoring technical research teams — whether in an academic lab, AI research organization, or industry settingStrong intuition for model behavior, alignment challenges, and post-training trade-offsExperience designing evaluation systems and thinking rigorously about what it means for a model to be readyAbility to communicate complex technical trade-offs clearly to both technical and non-technical audiencesA PhD or equivalent