Agentic media buying cannot scale without the right foundation. See how buyers and sellers get there on Databricks.

Summary

Media buying remains slowed by fragmented coordination across emails, spreadsheets, PDFs, and phone calls, leaving teams to research inventory, compare pricing, negotiate, and issue orders manually. The post presents an agentic workflow in which buyer and seller agents discover inventory, negotiate prices, and book deals through IAB Tech Lab’s AAMP standards, including AdCOM, OpenDirect, OpenRTB deals, and registry-based discovery and trust. Its Databricks implementation runs self-contained applications on Databricks Apps, with CrewAI agent crews using Databricks Foundation Model APIs and Claude models, Lakebase providing transactional Postgres state, and MLflow tracing capturing decisions and tool calls. The example processes a $200,000 Q3 Brand Launch across CTV and Linear TV with a $38 CPM ceiling, and is packaged as a Databricks Automation Bundle accelerator deployable with a single command.

Context

Media buying is described as a fragmented, largely manual process involving emails, spreadsheets, PDFs, phone calls, RFPs, rate cards, negotiations, and insertion orders. The lack of standardized inventory, audience, and trust mechanisms creates bespoke integrations and delays decisions while inventory, pricing, and audience signals change.

Approach / What changed

The implementation uses IAB Tech Lab AAMP standards for communication between independent buyer, seller, and registry applications. On Databricks, Databricks Apps host the parties, CrewAI organizes specialist agents, Databricks Foundation Model APIs and Claude models provide model access, Lakebase stores transactional state, built-in service principals and OAuth handle identity, and MLflow provides optional tracing. The accelerator deploys the apps, Lakebase instances, seeded data, and registry connections.

Takeaways

• IAB Tech Lab’s AAMP standards provide shared inventory and audience vocabulary, transaction protocols including OpenDirect and OpenRTB deals, and registry-based discovery and trust between compliant buyer and seller agents.
• The buyer agent crew uses a hierarchical structure: a Portfolio Manager splits budgets, Channel Specialists shop specific media, and tactical agents plan audiences and execute purchases.
• In the example, a $200,000 campaign allocates $120,000 to CTV and $80,000 to Linear TV; agents book only matching products priced at or below the $38 CPM ceiling.

Related reading