AI-Driven Multi-Agent System for Automated Social Media Marketing Campaigns
Megha Sindhu Indukuri,
Bhargavi Yerragolla,
Ramakrishna Vinayak Arigela,
MadhuBabu Gunji,
Venkata Siva Sri Dhanush Mada,
Shirin Bhanu Koduri
Multi-Agent Marketing System is put forth to automate the entire campaign lifecycle of digital strategic
development, personalization, content creation, poster generation, scheduling, and platform-specific
posting. Human coordination is predominantly needed for traditional marketing practices, resulting in
scalability challenges and varied delivery. Gemini generative models for ideation and text generation,
Stable Diffusion for poster generation using automated design, and APIs for distribution across platforms
such as Twitter, LinkedIn, and Telegram are used in the system. Each agent works autonomously with a
specific role, orchestrated by a central agent to make seamless execution and data exchange possible.
Experimental deployment indicates significant improvements in campaign execution time and accuracy
of tailored content, all with little human intervention. The architecture illustrates how Large Language
Models and Generative AI, when placed within an agentic framework, can provide scalable, consistent,
and adaptive marketing automation.
