I want to put a true “AI agent” at the heart of my social media and digital marketing workflow. Rather than a single-purpose bot, I’m after an autonomous agent that can fluidly switch between content creation, community engagement, analytics, and any other marketing task as the data demands. You’ll help me design, train, and deploy this agent so it can: • Pull fresh brand-relevant insights from my existing channels • Generate on-brand posts, captions, and visuals on the fly • Decide the ideal publish time and frequency • Learn from performance data and iterate its own strategy without manual prompts I’m platform-agnostic for now—Facebook, Instagram, Twitter, or any other network we agree makes sense—but the system must be modular so we can add or drop channels easily. Tech stack is your call: whether that’s a custom LLM pipeline, OpenAI tools, LangChain, AutoGen, or another framework. What matters is that the agent remains extensible; I should be able to plug in new APIs (for ads, email, SEO tasks) later with minimal re-engineering. Deliverables 1. A working AI agent deployed to a cloud environment I control (AWS, GCP, or Azure) 2. Clean, well-commented code plus setup documentation 3. A short loom or live walkthrough showing the agent publishing a post, gathering real-time feedback, and updating its next-step plan automatically Acceptance criteria • The agent can publish at least one autonomous post per day without human intervention after initial hand-off. • Performance metrics (reach, engagement, click-through) are logged and used for the next content decision. • I can trigger a manual override or tweak objectives via a simple dashboard or CLI. If you’ve built similar autonomous marketing agents before, show me! I’m ready to move quickly once I see a roadmap that proves this can be more than another rule-based scheduler.