We're looking full time AI researchers in the area of multi-agent systems, who will work with us either remotely or on-site at our NYC headquarters. Researchers will collaborate closely with our larger team on designing, developing and deploying real-world AI systems, from creating new models, algorithms and systems, to optimized deployment on multi-GPU systems at scale
The ideal candidate would have a Ph.D. in AI/ML/CS or related technical field, with a strong background in multi-agent systems and decision-making using foundation models. Strong interest in applying bleeding edge AI to real-world problems, and with research or industry experience
What you’ll do:
- Multi-agent and tool-using systems
- Sequential decision-making using language models including multi-step
- Reasoning and planning, and related training methods
- Continuous learning and self-learning
- Neuro-symbolic reasoning
- Contextual bandits, reinforcement learning and reward modeling
- Experience with classical planning techniques and AI agent frameworks is a plus
- Experience with large foundation models, including hands-on familiarity with different training modalities (CPT/DPT, SFT, instruction-tuning, alignment, ...) is a plus
Requirements:
- Graduate degree in Computer Science, Artificial Intelligence, Machine Learning, or related technical field (PhD students preferred)
- 5+ years of research or industry experience in one of the areas listed above
- Publications at strong AI/NLP conferences (Neurips, ICLR, ICML, ACL, NAACL, EMNLP, ...)
- Expertise in deep-learning frameworks including one or more of PyTorch, TensorFlow, Jax
The US base salary range for this role is $248,500 - $300,000, not inclusive of equity + benefits. Our salary ranges are determined by experience, skills, qualifications and location. The provided range on each posting details the minimum and maximum across all applicable locations in the United States. Your recruiter can share more information regarding our benefits package, equity as well as sales commissions (if applicable).
