AI Research Scientist (Datadog AI Research - DAIR)

Stack

Responsibilities

• ▹Conduct research in generative AI and machine learning, building specialized foundation models and trained agents for observability
• ▹Train multimodal models on large-scale, diverse telemetry data (metrics, logs, traces, topology, events) using distributed training infrastructure
• ▹Design and build simulated environments and RL training loops for on-policy agent training and evaluation
• ▹Collaborate with cross-functional teams (Product, Engineering) to integrate multimodal world modeling and autonomous agents into Datadog's products
• ▹Stay at the forefront of foundation models, world models, and RL-based agent research
• ▹Contribute to research publications, present at top-tier conferences (NeurIPS, ICLR, ICML), and help open-source key model artifacts and benchmarks

Requirements

• ▹PhD in Computer Science, Machine Learning, or a related field, with deep expertise in generative modeling, world models, AI agents, reinforcement learning, or multimodal learning (or equivalent experience)
• ▹Extensive experience designing and implementing deep learning models and agents, with a strong background in distributed training frameworks (e.g., DeepSpeed, Megatron-LM) and PyTorch
• ▹Track record of impactful publications at top-tier venues (NeurIPS, ICLR, ICML, TMLR)
• ▹Familiarity with efficient training, post-training, and inference techniques for large foundation models
• ▹Ability to explain complex models and research findings to both technical and non-technical audiences

Soft skills

About the company

Datadog AI Research (DAIR) tackles high-risk, high-reward research problems grounded in real-world challenges in cloud observability and security, building on Datadog's track record of AI-powered solutions (Bits AI, Bits Evolve, and its time series foundation model).