Using Agentic workflows to Optimize Media Proposals

Session details:

This paper presents an agentic workflow architecture designed to optimize media proposals in response to RFPs that combine structured campaign parameters with descriptive guidelines and qualitative objectives. As campaign briefs increasingly arrive through both traditional communication channels and emerging AI-driven protocols such as A2A and MCP, media organizations require intelligent systems capable of interpreting heterogeneous inputs and producing proposals that are simultaneously cost-efficient, goal-aligned, and revenue-maximizing. The proposed workflow employs specialized AI agents to analyze RFPs, identify missing details, generate clarification questions, and iteratively refine campaign requirements until a complete optimization context is achieved. Leveraging both quantitative audience-delivery metrics and qualitative campaign criteria, the AI agents select and configure media inventory products that minimize cost per outcome while filtering options to ensure alignment with the advertiser’s strategic intent.

The system further incorporates dynamic decision thresholds that determine when proposals may be automatically generated and approved versus when managerial oversight is required based on campaign size, inventory availability, and business rules. By orchestrating these functions within an agentic environment, media sellers gain the ability to respond rapidly with tailored, value-maximized proposals, while buyers receive recommendations that efficiently meet their performance goals and qualitative standards. This paper details the optimization algorithms, communication flows, and governance mechanisms that enable AI-driven proposal generation, demonstrating how agentic workflows simultaneously improve operational efficiency, enhance advertiser satisfaction, and elevate yield across media inventory portfolios.