Support
Agents that turn solved tickets and internal docs into help-centre articles, FAQs and searchable team knowledge.
12 published use cases
Every new Zendesk ticket gets a category, priority and sentiment tag plus a draft reply in an internal note before an agent opens it.
Cursor Agent writes and maintains interlinked markdown pages from user-supplied sources, summarizing and cross-referencing, with Obsidian used to browse.
Original post ↗An OpenClaw agent runs a YouTube summariser and a paper-condensing tool, saving Markdown into two indexed knowledge bases for videos and papers.
Original post ↗The author's wife plugged an MCP server into their household Wiki.js so an LLM can update the wiki.
Original post ↗Commenters use Codex for Pi-hole upgrades, market research, wiki building, file search, PowerPoint edits, light Excel work, scraping, and data entry.
Original post ↗Grok handled live chat, email, tickets, refunds, knowledge base, QA, reporting and escalations for an 8-person support company, per the author's account.
Original post ↗A team deployed an AI voice agent for after-hours calls, FAQs and order status, capturing structured call data for CRM and support systems.
Original post ↗An n8n workflow scrapes a website into a knowledge base, answers customer questions on WhatsApp and converts inquiries into booked calls.
Original post ↗The author set up a Synthflow receptionist that captured caller details and booked next-day appointments, and a Retell agent that answered clinic FAQs and flagged frustrated callers (vendor comparison).
Original post ↗An n8n agent looks up the knowledge base and order history, replies by email or escalates to Slack with a draft; billing and security go to humans.
Original post ↗Author runs a self-built autonomous agent on Codex, ChatGPT Plus, systemd and email that handles various company tasks, using a rolling context, diary and markdown wiki for memory.
Original post ↗An n8n and OpenAI agent answers WhatsApp questions from a knowledge base and escalates low-confidence cases to a human on Telegram with full context.
Original post ↗Yes. A common setup watches solved tickets for questions that keep coming up and drafts a new article or an update to an existing one. A support lead reviews it before it goes live.
Across the 12 published workflows on this page, the most used tools are n8n (4), Codex (2) and ChatGPT (1). Each use case lists every tool it needs and the role it plays.
Open any use case to see the tools, the step-by-step setup and example prompts you can copy. Start with a beginner workflow, run it on a small batch and review the output before you automate it fully.