SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 192,000 - 240,000 / annual
Datadog is seeking a Manager I for the Anomaly Detection team, which builds scalable, cost-effective systems to automatically detect anomalies in customer system telemetry. The team sits at the intersection of distributed systems and machine learning, applying these techniques at massive scale to solve real-world customer problems.
You will lead and develop an existing engineering team of 8 people, building trust, setting technical direction, and establishing high standards for ownership and execution. You'll own the roadmap and execution for key signal-finding systems, working closely with product management and applied scientists to build effective detection systems and evaluation mechanisms that ensure signal quality.
Key responsibilities include building and operating both traditional machine learning and LLM-based systems in production at scale, and driving the team's AI-native development practices by setting high standards for safety, validation, and increasing agent autonomy over time. You'll be comfortable making significant contributions to product strategy in ambiguous environments rather than simply executing a pre-defined roadmap.
Datadog operates as a hybrid workplace, valuing office culture for relationship-building, creativity, and collaboration while enabling work-life harmony.
REQUIREMENTS:
- Experienced engineering manager with a proven track record of shipping robust and scalable distributed systems with AI and/or ML components (distributed systems experience is a must-have)
- Proficiency with Java, Python, or Go, and experience working in platform-focused teams
- Strong cross-team collaboration and stakeholder management skills
- Comfortable making significant contributions to product strategy in ambiguous environments
- Direct experience building and training ML models in production environments (preferred)
- Familiarity with streaming technology such as Flink or Kafka (preferred)
