Senior AI Research Scientist – Enterprise Agents

Servicenow; Santa Clara (CA); On-site; USD 232,000 - 405,000; Full time

Benefits offered by this job

Health plans

401(k) Plan with company match

ESP P

Matching donations

Family leave programs

Job summary

ServiceNow in Santa Clara, CA is seeking a Staff Research Scientist to lead a major workstream in agent learning and recursive self-improvement. You will turn failures and successes into hypotheses, experiments, training signals, and deployable improvements to model weights and harness code.

This role blends scientific reasoning with training code, agent systems, and production constraints, requiring excellence in ML, distributed training, and collaboration with product and engineering teams to

Qualifications

• 10+ years of AI/ML research or engineering experience; PhD required.
• Proven ability to set technical direction and lead multiple high-impact projects across team boundaries.
• Strong foundations in ML, deep learning, reinforcement learning, and experimentation with large language or multimodal models.
• Advanced Python and PyTorch; experience modifying training code and data pipelines.
• Practical depth in agentic AI: tool use, planning, memory, retrieval, environments, or long-horizon execution.
• Experience with distributed training, rollout, or inference; modern post-training or serving stacks.
• Publications, shipped systems, patents, or open-source contributions.

Responsibilities

• Design and execute end-to-end research projects improving enterprise agents' capabilities across planning, reasoning, memory, tool use, retrieval, and coordination.
• Research post-training methods: continued pretraining, supervised fine-tuning, RL, DPO/GRPO, reward modeling, distillation.
• Develop harness-level optimization across prompts, tools, schemas, and memory management.
• Build improvement flywheels for data collection, trajectories, and signals; identify failure modes and interventions.
• Create realistic, stateful training environments and benchmarks with verifiers and graders.
• Run ablations and scaling experiments; reason about variance, distribution shift, and cost.
• Develop capabilities across modalities and multilingual settings.
• Build distributed pipelines for training, rollout, evaluation, and inference; optimize scale bottlenecks.
• Collaborate with researchers, engineering, and product to move validated methods to production.
• Communicate results through reviews, reports, publications, patents, and open-source contributions.

Skills

AI/ML research

Python

PyTorch

Distributed training

Research leadership

Agentic AI

Evaluation design

System design

Multimodal/ML models

Education

PhD or equivalent advanced degree

Tools

Research infrastructure

Experiment pipelines

Job description

ServiceNow in Santa Clara, CA is seeking a Staff Research Scientist to lead a major workstream in agent learning and recursive self-improvement. You will turn failures and successes into hypotheses, experiments, training signals, and deployable improvements to model weights and harness code.

This role blends scientific reasoning with training code, agent systems, and production constraints, requiring excellence in ML, distributed training, and collaboration with product and engineering teams to