The management of social media for brands and individuals is a critical but highly inefficient process. Professionals rely on a fragmented suite of tools for content creation, formatting, and distribution, a disconnection that hinders the ability to execute a cohesive strategy. This paper introduces Phase 2 of PostHero, an AI-powered, multi-platform social media manager designed to solve these challenges at scale. Building upon an initial single-platform proof-of-concept, the system's core contribution is a novel, stateful agentic workflow built with LangGraph, which dynamically manages the entire content lifecycle across both Instagram and Facebook simultaneously. This architecture integrates generative AI for text and image creation (via Groq and Pollinations AI), automated cross-platform API routing via FastAPI, and real-time market trend ingestion powered by Pytrends. Furthermore, it introduces a native, zero-dependency asynchronous scheduling engine to manage concurrent API publications without external message brokers. By creating a proactive, data-driven workflow, the system automatically aligns generated multi-platform content with current public interest. Results demonstrate that unifying these core functions through a multi-agent AI significantly streamlines the multi-platform workflow, maintaining a rapid time-to-publish efficiency of under 3 minutes while successfully automating distribution and formatting across distinct social network APIs.
Barbhai et al. (Wed,) studied this question.