In many business operations, two parties need to communicate regularly over email β for example, to exchange purchase orders, confirm deliveries, or request updates. However, in some workflows, it's essential to keep these parties separate: they shouldn't have direct contact or see each other's addresses.
To solve this, we designed an AI-powered email intermediary β an automation agent that acts as the bridge between both sides, handling all message routing, context management, and thread continuity automatically.
The Challenge
Each exchange follows a long, multi-step email chain:
- One side sends a new order or request (identified by a unique PO number).
- The other side replies, sometimes multiple times, attaching documents or follow-ups.
- The communication must always flow through an intermediate address, preserving thread history while keeping both parties separate.
Manually managing these threads was error-prone and time-consuming. The goal was to build an automation that:
- Detects and classifies new incoming emails.
- Extracts and cleans the main content (removing signatures, greetings, and previous quotes).
- Creates a draft reply to the opposite party, keeping the conversation within the same thread.
- Tracks each conversation in a database to maintain continuity.
- Sends reminders if no response is received within a defined period.
The Solution
We implemented the system using Zapier, combining:
- Gmail for email reception, drafting, and threading.
- Google Sheets as a lightweight database to track active conversations and thread IDs.
- OpenAI for intelligent text parsing β cleaning, extracting relevant content, and maintaining professional tone consistency.
The workflow logic automatically distinguishes:
- Whether the message is incoming from Side A or Side B.
- Whether it's part of an existing thread or a new one.
Each new chain is assigned a unique identifier, and the automation dynamically looks up existing records before deciding whether to create a new entry or update an ongoing conversation.
This ensures that every draft reply is attached to the correct existing thread, maintaining context and clarity across both sides.
Outcome
The result is a virtual email coordinator β a reliable, automated intermediary that keeps multi-party communication organized, consistent, and efficient.
By integrating intelligent text handling (OpenAI) with no-code workflow automation (Zapier), we were able to deliver a system that:
- Reduces manual forwarding and drafting time
- Keeps every conversation properly threaded
- Eliminates the risk of cross-contact between both sides
- Can evolve into a fully autonomous agent after manual validation
This AI-powered email intermediary demonstrates how intelligent automation can handle complex, multi-party communication workflows while maintaining security, context, and professional standards β all without human intervention.