AI Product 2026 Internship
•Brain Co.· 8h
Data ScienceFull-timeAmericas
8h ago
Phoebe·about 1 year ago
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.
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.
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.
An AI-native engineer: You're already using AI as an essential part of your workflow. You've discovered creative ways to leverage AI for rapid prototyping, debugging complex issues, or exploring architectural decisions. You see AI tools as force multipliers, not crutches.
A production ML expert: You’ve built ML systems at scale with strong Python skills and distributed systems understanding. You bridge the gap between research and production with robust engineering practices.
An LLM practitioner: You've built real systems with foundation models—RAG implementations, agent frameworks, or custom applications. You understand agent engineering deeply and have opinions on orchestration patterns.
A multi-disciplinary thinker: You combine expertise across NLP, information retrieval, and knowledge representation. You've built search systems or knowledge graphs and naturally think about structuring and querying information.
A root cause detective: You understand causal inference beyond correlation. Whether through graphical models, counterfactual reasoning, or temporal analysis, you can design systems that uncover true root causes.
A pragmatic builder: You balance cutting-edge ML with practical constraints, thriving in ambiguity while driving projects from prototype to production. You know when simple beats complex.
Note: This is an ambitious role—we're looking for exceptional talent in key areas rather than checking every box. If you excel in several of these areas and are excited to grow in others, we'd love to hear from you.
Experience working with telemetry data—turning logs, metrics, and traces into actionable insights
Track record of optimising models through fine-tuning, distillation, or human-in-the-loop techniques
Contributions to the ML community through research papers, conference talks, or open source projects
Prior work on agentic systems, whether in academia or industry
This is your chance to work on genuinely hard problems at the intersection of AI and production systems. If you're excited about shipping autonomous systems into mission-critical infrastructure, let's talk.
Impact: Join as an early employee with direct influence on product and company direction
Compensation: Competitive salary and compelling equity as an early employee
Culture: Remote or hybrid for those near our London office (2-3 days/week), quarterly European off-sites
Building in the open: All employees are owners and company builders—you'll receive investor memos and updates, with full transparency into our strategy and metrics
AI experimentation budget: Dedicated personal budget for AI tooling to stay at the cutting edge
Home office setup: Generous allowance to create your ideal work environment
Learning culture: Team that prioritises knowledge sharing through regular tech talks, pair programming, and conference attendance