- Experience
- 0+ yrs
- Location
- remote
- Job type
- Full-time
Required skills
- ml
- ai
- software engineering
- causal analysis
- information retrieval
- model distillation
- logs
- metrics
- traces
- multi-agent coordination
- decision-making under uncertainty
- low-latency
- high-throughput
- caching
- routing
- multi-modal analysis
- temporal patterns
- causal relationships
- evaluation infrastructure
- automated evaluation
- drift detection
- fine-tuning
- rag architectures
- synthetic data generation
About the role
Phoebe is a VC-funded startup transforming how engineering teams improve the reliability of complex systems. Founded by ex-Google, Meta, Amazon, and Stripe engineers, we're building AI agents that detect, diagnose and fix software problems to reduce errors and downtime.
The Challenge
Modern software systems generate millions of events per second across logs, metrics, and traces—but finding the root cause of an incident can still take hours of manual investigation. As an ML engineer, you'll build systems that analyse massive volumes of observability and business data, perform multi-step reasoning, and deliver actionable insights in real-time. You'll pioneer approaches that blend causal analysis, information retrieval, and model distillation to help our agents understand not just what went wrong, but why.
What You'll Do
Design agentic architectures: Build reasoning workflows that combine foundation models with custom tools for autonomous root cause analysis. Develop novel approaches to multi-agent coordination and decision-making under uncertainty.
Optimise for production scale: Deploy ML systems for low-latency, high-throughput environments. Implement caching and routing strategies to handle millions of events at sub-second response times.
Pioneer multi-modal analysis: Create systems that jointly analyse logs, metrics, traces, alerts and business data. Build representations that capture temporal patterns and causal relationships.
Build evaluation infrastructure: Create testing and monitoring systems for agent performance, including automated evaluation pipelines and drift detection.
Drive ML innovation: Experiment with fine-tuning, RAG architectures, and synthetic data generation. Translate research breakthroughs into production features.
About Phoebe
Phoebe is a VC-funded startup transforming how engineering teams improve the reliability of complex systems. Founded by ex-Google, Meta, Amazon, and Stripe engineers, we're building AI agents that detect, diagnose and fix software problems to reduce errors and downtime.
This page is fully interactive when JavaScript is enabled. Please enable JavaScript to apply or browse related roles.