SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Datadog AI Research (DAIR) is seeking an AI Research Scientist to conduct fundamental research on world models and trained agents for cloud observability and security. You will partner with Research Engineers and cross-functional product and engineering teams to translate cutting-edge AI research into production systems.
The role focuses on two core research areas: (1) World Models for Observability—training multimodal foundation models that learn the joint dynamics of distributed systems across metrics, traces, logs, topology, and events to power forecasting, anomaly detection, root cause analysis, and counterfactual simulation; and (2) Trained Agents for Observability—post-training models to operate autonomously in SRE incident response, code repair, security response, and infrastructure optimization.
Key responsibilities include conducting research in generative AI and machine learning; training multimodal models on large-scale telemetry data using distributed training infrastructure; designing and building simulated environments and RL training loops for agent training and evaluation; collaborating with product and engineering teams to integrate capabilities into Datadog's products; staying current with foundation models, world models, and RL-based agent research; and contributing to research publications and top-tier conference presentations (NeurIPS, ICLR, ICML).
You should hold a PhD in Computer Science, Machine Learning, or related field with deep expertise in generative modeling, world models, AI agents, reinforcement learning, or multimodal learning. You need extensive experience designing and implementing deep learning models with strong knowledge of distributed training frameworks (DeepSpeed, Megatron-LM) and PyTorch. A track record of impactful publications at top-tier venues is required, along with familiarity in efficient training, post-training, and inference techniques for large foundation models. You should communicate complex research findings clearly to both technical and non-technical audiences.
Bonus qualifications include experience bridging research and real-world product applications with foundation models or RL-trained agents, passion for customer impact and scalable AI deployment, production data pipeline experience, and hands-on GPU programming and optimization (CUDA).
Datadog offers competitive global benefits, new hire stock equity (RSUs) and ESPP, collaboration opportunities across NYC and Paris offices, conference attendance and presentation opportunities, mentorship programs, and an inclusive company culture with employee resource groups.
