The Prospecting Science program runs Al agents that engage prospects at enterprise scale across 4 regions, 12 languages, and all customer segments. Behind every agent interaction is a prompt architecture, a signal workflow, and a quality gate and someone has to build, tune, and maintain them all. This is the most hands-on-keyboard role in the program. You will spend the majority of your time building: designing prompts, configuring agent behaviors, connecting data signals to message logic, and shipping self-serve Al workflows that sellers use daily. This position is an individual contributor role reporting to the Sr Director, Growth Global Prospecting and Inbound Sales. Responsibility Build and maintain Al agent configurations: prompt architectures, context injection logic, tone calibration, and response validation rules Design signal-to-message workflows end to end: from enrichment (firmographic, technographic, intent, behavioral) through message generation and quality scoring Own the business requirements document (BRD) for use cases assigned by the Growth Al Lead translate business intent into buildable specifications Ship self-serve Al workflows using tools like Clay, enrichment APIs, and prompt orchestration platforms no engineering sprint required Build and maintain technical “experiment-in-a-box” documentation: prompt libraries, configuration specs, agent behavior guides, and testing protocols Operate with quality obsession: test every configuration against real prospect data before production deployment Instrument quality telemetry: build monitoring for output quality, response rates, and failure modes so the team can iterate from data Collaborate with DTS on complex, multi-system builds that require platform-level infrastructure Maintain the shared roadmap with build status, shipping dates, and technical dependencies