AI Research Scientist: Multimodal World Models

Datadog; New York (NY); On-site; USD 320,000 - 400,000; Full time

Benefits offered by this job

Competitive benefits

RSUs & ESPP

Office in NYC & Paris

Conference opportunities

Mentor program

Employee resource groups

Job summary

Datadog researches foundational AI for observability, partnering with research engineers to translate breakthroughs into products.

You will advance multimodal world models and autonomous agents, trainer pipelines on large telemetry data, and collaborate with Product and Engineering to integrate capabilities into Datadog's suite.

The role emphasizes publishing at leading venues and contributing to open-source artifacts while benefiting from Datadog's global growth and inclusive culture.

Qualifications

• PhD in CS, ML, or related field with expertise in generative modeling and RL.
• Experience with distributed training frameworks and ML libraries.
• Strong publication record in NeurIPS/ICML/ICLR.

Responsibilities

• Conduct research in generative AI and machine learning, building specialized foundation models and trained agents for observability
• Train multimodal models on large-scale telemetry data using distributed training infrastructure
• Collaborate with cross-functional teams to integrate capabilities like multimodal world modeling and autonomous agents into Datadog's products

Skills

PhD in CS/ML

Deep learning models

Multimodal learning

Reinforcement learning

Distributed training

PyTorch

Publications in top venues

Education

PhD in Computer Science / Machine Learning

Tools

DeepSpeed

Megatron-LM

CUDA

PyTorch

Job description

Datadog researches foundational AI for observability, partnering with research engineers to translate breakthroughs into products.

You will advance multimodal world models and autonomous agents, trainer pipelines on large telemetry data, and collaborate with Product and Engineering to integrate capabilities into Datadog's suite.

The role emphasizes publishing at leading venues and contributing to open-source artifacts while benefiting from Datadog's global growth and inclusive culture.