Lago, the open-source billing platform backed by $22M in Series A funding, is hiring a Product Engineer to transform its internal AI agents from monolithic skills into a scalable multi-agent platform. The company powers usage-based pricing for major clients like PayPal, CoreWeave, and Mistral AI, handling real-time metering and complex billing workflows.

Currently running on Claude Code, these agents automate lead qualification, prospect research, and customer monitoring but operate as single mega-prompts controlled by one person. The new hire will decompose these into composable agents with clear inputs and outputs, making the system multi-user with shared state management and different permission levels. The role involves building self-improvement loops where feedback signals automatically calibrate agent performance.

The position offers direct collaboration with founders and starts from working prototypes with months of real calibration data. Success metrics include decomposing the first skill within three months, launching the self-improvement loop by six months, and establishing a platform where new agents can be spun up in days rather than weeks within a year.