Research AI Director
Hybrid · Donostia / San Sebastián, Basque Country, Spain
Job Summary
Research AI Director leads a senior, multidisciplinary AI research organization to shape and execute Multiverse Computing's AI research strategy, driving LLM and AI development from research to production. Responsibilities include defining a multi-year AI research roadmap (foundation models, model compression, alignment, applied AI), translating scientific vision into deliverable programs, overseeing end-to-end LLM development lifecycle (pretraining, fine-tuning, RLHF/DPO, evaluation, deployment), managing and scaling a diverse team of researchers, ML engineers, software engineers, and applied scientists, establishing KPIs and publication targets, coordinating cross-functional collaboration (Product, Engineering, Business Development), and representing the company at top-tier conferences and advisory forums. Qualifications emphasize a PhD with a strong publication record, 8+ years post-doc/industry experience, deep LLM expertise, experience bridging academia and industry, distributed-training proficiency, and excellent communication and leadership skills. Perks include an indefinite contract, visa sponsorship where applicable, relocation package, and hybrid/flexible working arrangements.
Required Qualifications
- PhD in Computer Science, Machine Learning, Mathematics, or a closely related field with a strong peer-reviewed publication record in AI/ML at top venues (NeurIPS, ICML, ACL, ICLR, EMNLP, or equivalent)
- 8+ years of post-doctoral or industry research experience
- At least 2 years in a senior AI research leadership or management role
- Deep, proven expertise in large language model (LLM) development: pretraining, fine-tuning, RLHF/DPO alignment, evaluation, and deployment at scale
- Demonstrated track record of managing and scaling AI research or engineering teams in both academic and commercial/industrial environments
- Hands-on experience with distributed training infrastructure: multi-GPU/multi-node clusters, Megatron-LM, DeepSpeed, or FSDP, and performance profiling at scale
- Ability to operate with academic rigour (publications, reproducibility) and industrial discipline (delivery timelines, product integration)
- Strong strategic thinking with ability to set direction and communicate complex research to non-specialist audiences
- Excellent interpersonal and stakeholder management skills
Apply with one swipe on Sorce. We auto-fill applications and apply on your behalf — no cover letters, no 40-minute forms.
Hiring someone like this?
Get your role in front of qualified candidates on Sorce.