Databricks AI Tackles Ad Strategy Gap

Advertisers and agencies often struggle to translate high-level campaign strategies into actionable data segments. This disconnect leads to diluted messaging and missed opportunities, as the teams closest to strategy are furthest from the data, and vice versa. Databricks aims to resolve this with a new AI-powered audience generation solution built on its Data Intelligence Platform.

StartupHub data

Companies working on this

Profiles of the companies named in this story, with funding and a one-liner from our database.

Databricks

$190.0B

A unified data analytics and AI platform built on the lakehouse architecture.

This approach leverages agentic AI to allow planners to describe target audiences in natural language. The system then translates these prompts into precise data segments, drastically cutting down the weeks-long iteration cycles traditionally required. This marks a significant step in applying a multi-agent approach to audience intelligence, according to Databricks.

Bridging Strategy and Data Execution

The core challenge lies in the translation from a campaign brief to data-driven execution. Strategy dilution occurs when nuances are lost in translation, while incomplete strategies emerge when data insights fail to loop back to the business. Data blind spots also persist, as even vast datasets contain complexities that human analysts may overlook.

Databricks' solution uses its latest advancements in agentic AI to create a seamless bridge. The platform ingests first-party and partner data, curates it into unified population attributes, and then utilizes specialized agents for audience generation and insight discovery.

Automated Audience Generation and Discovery

At the forefront is Databricks Genie, a component that allows advertisers to define audiences using natural language. For example, a planner for a luxury travel brand can request "affluent travelers aged 35-54 who frequently book premium experiences." Genie translates this into governed SQL queries against millions of records in seconds, providing transparency into the logic used. This democratizes data access, shifting audience creation from a technical task to a strategic conversation.

The platform's Affinity Agent goes a step further by identifying non-obvious patterns within these defined audiences. By running statistically validated lift calculations against massive datasets, it can uncover surprising correlations, such as luxury travelers over-indexing on cryptocurrency investments or showing a strong affinity for wellness content. These statistically validated discoveries enrich campaign strategy and media placement decisions.

Orchestrating these agents is Agent Bricks, a supervisor that connects strategic intent with data execution. This unified AI orchestration collapses traditional, lengthy iteration cycles into a single, conversational workflow, enabling real-time responses to market shifts and client requests. This innovative framework streamlines AI agent deployment.

The Power of Orchestrated Intelligence

The entire system is designed for speed and accuracy. Data sources are ingested and curated into a unified table within Unity Catalog. The Genie Space translates natural language requests, while the Affinity Agent analyzes audience behaviors. Finally, Agent Bricks orchestrates the multi-agent system, routing requests and ensuring seamless operation.

Audiences generated are saved as tables in Unity Catalog and can be activated across downstream channels like DSPs, email platforms, and social media. This process not only accelerates campaign activation but also embeds strategic intent directly into audience generation, leading to demonstrably better targeting and performance.

The platform's architecture is built for production readiness, ensuring that the expertise encoded in the data layer and metadata is continuously improved and scaled across the organization. This approach transforms audience creation from a bottleneck into a significant competitive advantage.

This new capability builds upon Databricks' ongoing efforts to enhance AI development and deployment, including innovations like Databricks Genie and the framework for Agent Bricks, further exemplified by their work on models like Qwen3.