What if the data owner, media buyer, ad platform, and normalization and delivery layer could each just do their best work, while an agent handles the handoffs - and valuable real-time data went from raw to immediately usable? This session unpacks that model: The Weather Company builds agentic skills around forecasting and weather triggers, Narrative brings the connective infrastructure, and an agent sits between every party, turning what each needs into what the others can use, live, with no custom integration.
Every industry generates real-time information that loses value the moment it goes stale, and turning it into revenue usually takes four parties, each holding something the others can't replicate. In this case, that's The Weather Company, Narrative, the activation platforms, and the media buyer. Today, that means hand-building a connection for every pair, and starting over each time. This session shows a different model, where skills are the unit of composition: The Weather Company packages its forecasting and weather-trigger logic as agentic skills, Narrative brings its own skills for normalization and delivery, the activation platforms plug in through their existing APIs and contracts, and a media buyer states an objective plainly to an agent, which composes the right combination across all four, in whatever harness they already run. Attendees leave with a concrete look at how valuable real-time data goes from raw to usable in practice, without anyone rebuilding their stack.
Every industry generates real-time information that loses value the moment it goes stale, and turning it into revenue usually takes four parties, each holding something the others can't replicate. In this case, that's The Weather Company, Narrative, the activation platforms, and the media buyer. Today, that means hand-building a connection for every pair, and starting over each time. This session shows a different model, where skills are the unit of composition: The Weather Company packages its forecasting and weather-trigger logic as agentic skills, Narrative brings its own skills for normalization and delivery, the activation platforms plug in through their existing APIs and contracts, and a media buyer states an objective plainly to an agent, which composes the right combination across all four, in whatever harness they already run. Attendees leave with a concrete look at how valuable real-time data goes from raw to usable in practice, without anyone rebuilding their stack.
Speakers
Key Takeaways
- Turning real-time data into revenue takes four parties - the data owner, the normalization layer, the activation platforms, and the media buyer - and none gets there alone.
- Skills, not integrations, are becoming the unit of composition - letting separate parties combine expertise without either side adopting the other's tooling.
- Normalization is what makes that composability possible - once real-time data is normalized, it activates consistently on any downstream platform.
- What this means for media buyers: state an objective to an agent, and get a strategy composed across multiple parties, no new tooling required.
- A framework any business can use to evaluate build-vs-plug-in tradeoffs when adding real-time data to an existing ad stack.
Presented by
