Madison World plans to roll out the first version of its proprietary artificial intelligence platform, Madison M-BrAIn, in early April as the independent agency group accelerates its transition to AI-driven planning and product-driven media services.
Ajit Varghese, partner and group CEO of Madison Media, OOH and Hiveminds, said the platform, which could require an investment of around $1 million, is designed to transform the way the agency approaches its media strategy by combining internal knowledge with external data sources and advanced AI models.
Varghese said the effort is part of a broader overhaul of the agency’s planning system as Madison transitions from a traditional media planning framework to what it calls the Growth Planning System (GPS).
“Traditionally, agencies built frameworks around media planning and allocation. We’re redesigning that structure into what we call a growth planning system,” Varghese said.
Advertising effectiveness is typically measured by factors such as attention, memory, and response. However, Varghese believes that planning in a complex market like India needs to start early in the decision-making process.
He said that before recommending a media strategy, planners first need to understand the underlying dynamics of the market, including consumer barriers, purchase triggers, and brand challenges. Once these factors are identified, agencies can develop a clearer growth agenda for their clients and design a media strategy that more closely aligns with their business goals.
To operationalize this approach, Madison developed Madison M-BrAIn, a human and AI cognitive ecosystem that serves as a central intelligence hub for planners. The system aggregates unique insights using information from external sources and large-scale language models, allowing planners to gain deeper market intelligence before building campaign strategies.
Varghese said the goal is to democratize strategic knowledge across the organization.
“Until now, this level of understanding has been primarily available only to senior leaders and experienced strategists. With Madison M-BrAIn, even junior planners should be able to access the same information and return to clients with a more informed perspective,” he said.
The agency has already rolled out the new planning philosophy internally and completed a three-month test of its AI platform. Varghese said early trials have shown promising results in terms of system learning and performance.
While initial versions of the platform relied on the global large-scale language model, Madison is now developing its own small-scale language model (SLM) that will serve as the core of the Madison M-BrAIn ecosystem.
“The SLM can read the global LLM, but the LLM cannot read the SLM,” Varghese explained. “This ensures that all the intelligence we build remains within the Madison ecosystem and continues to strengthen our unique knowledge base.”
The company plans to release the first version of Madison M-BrAIn in early April, with a more sophisticated version expected by the end of June. Following the initial release, the platform will gradually integrate additional external data streams and APIs, including consumer insights platforms, social listening tools, and client datasets.
Over time, these integrations are expected to enhance the system’s learning capabilities and enable it to generate more sophisticated strategic recommendations.
Madison M-BrAIn is currently being deployed internally, but Varghese said the platform could eventually evolve into a product for external users as the system matures.
“At this time, our focus is to stabilize and strengthen Madison M-BrAIn internally. However, over time, this may become an externally licensed product,” he said.
The investment in Madison M-BrAIn is part of a broader technology transformation underway at the agency group. In addition to the AI platform, Madison is building something called the Catalyst operating system. It aims to integrate the company’s digital infrastructure, operational processes and product platform into a unified ecosystem.
Varghese said the broader technology stack could require an additional $1 million to $1.5 million investment over time, but the spending would be phased in rather than being deployed all at once.
“We review our progress every three months and prioritize the most important features first,” he said.
Varghese added that in a rapidly evolving technology environment, agencies cannot afford to delay investing in AI and data-driven capabilities.
“Speed is critical in a technology environment. If you take too long to adapt, you risk becoming irrelevant,” he says.
Madison expects its extensive AI and operating ecosystem to be fully operational within 12 to 18 months, enabling it to combine human strategy and machine intelligence in future customer engagements through Madison M-BrAIn.
first publication date
