Automatically classify support messages and generate instant AI-written replies that match your brand tone.
A complete AI powered support system that reads customer messages, understands intent and sentiment, drafts polished replies, and organises every ticket inside Attio without manual effort. Users get faster response times, more consistent communication, lower support workload and a structured support workflow that scales easily as message volume grows.
Set up an Attio workspace and create a Support Tickets list with fields like customer name, email, channel, message, intent, sentiment, AI reply, status and assignee.
Configure your support inbox or contact form so that every new customer message is forwarded into Zapier as a trigger event with the full message body and sender details.
In Zapier, add a step that sends each incoming message to Relevance AI so it can classify the ticket intent, detect sentiment and extract key entities such as product, feature and urgency.
Map Relevance AI outputs into clean fields like intent, sentiment, topic and urgency, then pass those fields along with the original message into the next Zapier action.
Add a Zapier step that sends the structured data and raw message into ChatGPT with a prompt that tells it to write a clear, accurate support reply in your brand tone using the detected intent and sentiment.
Capture ChatGPT's response and store it as the AI reply draft, then create or update the ticket in Attio with all parsed fields plus the suggested reply and set the status to Needs review or Ready to send.
Configure an internal workflow where your team reviews AI drafted replies in Attio for edge cases or sensitive topics and either approves them as is or edits before sending.
Add an optional Zapier step that, once a reply is approved in Attio, sends the AI generated response back to the customer from your support inbox and updates the ticket status to Resolved or Waiting on customer.
Set up filters and views in Attio so support leads can quickly see tickets by intent, sentiment, urgency and response time and adjust prompts or classification rules when patterns emerge.
Review a small batch of resolved tickets each week, feed examples into ChatGPT to refine your base prompt and update your Relevance AI configuration so the system becomes more accurate and on brand over time.
Your AI assistant for everyday tasks, creative ideas and advanced productivity.
Build and deploy custom AI agents to automate business workflows.
CRM re-engineered for high-velocity revenue teams.
Automate your apps and workflows seamlessly with no code required.
This blueprint creates an automated customer support system that reads every incoming message, understands what the customer needs, drafts a complete reply, and stores everything in a structured support pipeline. The workflow uses Relevance AI to interpret messages, ChatGPT to generate high quality responses, Attio to track all support activity, and Zapier to connect the entire process so it runs automatically.
The workflow starts when a customer sends a message through your support inbox or contact form. Instead of manually reading each message, Zapier captures it instantly and sends the full text to Relevance AI. Relevance AI then analyses the content and extracts the important details such as intent, sentiment, topic, request type, and urgency. This gives you a clean understanding of the customer’s issue even before writing a reply.
After Relevance AI finishes processing the message, Zapier sends both the structured data and the raw message content to ChatGPT. With a clear prompt that includes your tone guidelines, brand voice, product context, and the extracted intent and sentiment, ChatGPT produces a complete support reply that is accurate, empathetic and aligned with your communication style. This eliminates the inconsistency that normally comes from having multiple support agents responding differently.
Zapier then posts everything into Attio. Each ticket gets a full profile that includes the original message, the extracted intent and sentiment, the AI drafted reply, and the ticket status. Attio becomes the central support dashboard where your team can view open tickets, approve replies, adjust messages for sensitive cases and resolve issues efficiently. This replaces scattered emails and messy inboxes with a clean, organised workflow.
The system can be configured in two modes. In the standard mode, AI generated replies are sent to Attio for review before a human sends them manually. This helps maintain control in early stages. In the advanced mode, you can allow Zapier to send approved responses directly back to customers once the ticket status is set to Ready to send. This creates a fully automated loop that handles most support messages without human intervention.
The blueprint continues improving over time. Every resolved ticket creates more data that you can feed back into ChatGPT to refine prompts and improve response quality. You can also update Relevance AI extraction rules when new product features or common issues appear. Because everything is structured in Attio, you can quickly identify repeated issues, common sentiment patterns and trending support topics.
The final result is an AI powered support system that responds to customers quickly, keeps tone and quality consistent and frees your team from repetitive message handling. It reduces response times, lowers support workload and creates a smoother experience for customers. All of this is achieved with four tools that work together seamlessly and require minimal maintenance once set up.
Share it with your network