AI Automation Specialist: Build Context-Aware (Triggerless) Comment Responder for Instagram & TikTok

We are looking for an experienced automation and AI integration specialist to build an intelligent, context-aware comment response and DM follow-up system for our social media accounts (Instagram and TikTok). Unlike conventional rule-based automations that rely on specific keywords (e.g., commenting "INFO" or "LINK"), this system must evaluate incoming comments organically using Large Language Models (LLMs). The core objective is to identify when a user is asking a wellness or health-related question, generate an empathetic, contextually appropriate public response with non-medical disclaimers, and optionally initiate a private direct message sequence. We are platform-agnostic: you may build this using ManyChat, Make.com, Zapier, custom Webhooks / Meta Graph & TikTok APIs, or specialized comment-AI tools (e.g., Rypl, Inrō, Chatfuel)—as long as it complies strictly with official platform policies and avoids shadowban/rate-limit risks. Key Responsibilities & System Requirements Context-Aware Topic Classification Listen to all incoming post and video comments without relying on static keyword triggers. Use an LLM step (e.g., OpenAI GPT-4o / Claude) to classify intent (Health/Wellness Inquiry vs. Spam vs. General Praise vs. Unrelated). Automated Dynamic Response Generation Draft varied, non-robotic public replies that validate the user's inquiry naturally. Automatically append required educational/non-medical disclaimers. Direct users to a secure channel (e.g., DM or link-in-bio resource) for further details. Safety Guardrails & Triage Implement strict negative prompts and keyword filters: any signs of medical emergencies, acute distress, or self-harm must immediately bypass standard replies and post standardized crisis intervention guidance (or flag for immediate human review). Include a review queue or high-risk escalation webhook (e.g., alert sent to Slack, Telegram, or email). Multi-Platform Deployment & Compliance Primary: Instagram (Reels & Feed Posts). Secondary: TikTok (Video comments via official TikTok API or supported automation middleware). Zero scraping or unauthorized browser automation; 100% official API integrations only. Deliverables Architecture & Workflow Setup: Fully functional integration connected to our social media profiles via your recommended stack (ManyChat, Make.com, Zapier, or direct API). System Prompt & Guardrail Engineering: Tested system prompts with temperature controls, classification schemas, and safety boundaries tailored to health/wellness content. Testing & Edge-Case Validation: Recorded test runs demonstrating response variations across at least 15 sample comments (general questions, severe health concerns, spam, edge cases). Documentation & Handoff: A clear 1–2 page operating guide (or Loom walkthrough) detailing how to adjust prompts, view logs, pause automations, and manage API keys. Required Skills & Qualifications Proven track record building social media automations using official Meta Graph API and TikTok APIs / official partners. Deep expertise in workflow platforms (Make.com, Zapier, ManyChat, or custom Python/Node.js webhooks). Practical experience in Prompt Engineering & LLM integration (OpenAI API, Anthropic, or LangChain) for content classification and real-time generation. Familiarity with social media compliance, rate limits, and anti-spam best practices. Experience in regulated or high-sensitivity niches (health, wellness, legal, or finance) is a major plus.