Interfere is building an AI-powered platform to detect and diagnose product failures in real time. This role owns the intelligence layer: building agents, LLM pipelines, evals, and context engineering systems that find bugs before users hit them. You'll work at the intersection of research and production, shipping systems that run at scale against real codebases.
Responsibilities
Build agents that reason over codebases, traces, and logs to detect and explain bugs.
Develop LLM pipelines for triage, root-cause analysis, and automated fix proposals.
Design detection systems for anomalies in product behavior and user experience.
Create evals, datasets, observability, and feedback loops to measure and improve system performance.
Implement context engineering for retrieval, indexing, and context construction.
Requirements
Experience shipping ML/LLM systems to production.
Strong ability to move between research and engineering.
Proficiency with agent architectures, tool use, and multi-step reasoning.
Treat evals as first-class engineering problems.
Can own ambiguous AI problems end-to-end.
Nice to Have
Background in code understanding or program analysis.
Interfere provides a production visibility platform that monitors real user behavior in applications to identify, understand, and automatically resolve bugs and unseen issues. It is designed to help engineering, product, and design teams improve efficiency by reducing time spent on routine debugging.