Senior Applied Scientist: Real-Time Anomaly ML for Security

Datadog; Paris; Hybrid; EUR 120,000 - 180,000; Full time

Benefits offered by this job

RSUs

ESPP

Professional development

Conferences & publishing

Mentor program

Inclusive culture

Global benefits

Job summary

Datadog is hiring a Senior Applied Scientist to build and optimize AI-based anomaly detection models for security products. You will design architectures, train at scale, and optimize implementations to run efficiently on production traffic, balancing speed and cost on large datasets.

You will lead data pipelines, integrate models with engineering teams, and push for interpretable, reliable outputs, working in a hybrid office environment in Paris.

Qualifications

• Hands-on experience training and fine-tuning models at scale.
• Deploying models into production with throughput and cost constraints.
• Working knowledge of GPUs and optimizing for hardware.
• Strong applied-mathematics fundamentals.
• Ability to explain trade-offs to engineers and product partners.

Responsibilities

• Design and build custom mid-size models for high-throughput stream processing and train them at scale.
• Optimize these models across mathematics and systems, aligning with hardware constraints.
• Build training data pipelines from raw streams to training-ready datasets.
• Collaborate with engineering to integrate models into production focusing on GPU utilization and latency per record.
• Plan the roadmap of model and system improvements based on product needs.
• Build an agentic layer to analyze, validate, and act on outputs of the models.
• Create lightweight interpretability tools to explain model behavior to users.
• Maintain and monitor models, services, and infrastructure and participate in on-call rotation.

Skills

Model training at scale

GPU optimization

Applied mathematics

Communication with engineers

Software design & implementation

Education

BS/MS/PhD in CS/Engineering/ML/Math

Tools

GPUs

Job description

Datadog is hiring a Senior Applied Scientist to build and optimize AI-based anomaly detection models for security products. You will design architectures, train at scale, and optimize implementations to run efficiently on production traffic, balancing speed and cost on large datasets.

You will lead data pipelines, integrate models with engineering teams, and push for interpretable, reliable outputs, working in a hybrid office environment in Paris.