How to build a Proposal Reference Agent

This workflow streamlines the process of researching past proposals and sending a summary to a specified recipient. Here’s how it works:

1. User Inputs

• Proposal Topic Input:
  The user enters the topic for a new proposal (e.g., “International airport improvement project with sustainability focus”).

• Recipient Email Input:
  The user provides the email address of the person who should receive the summary.

2. Knowledge Base Search

• The workflow searches a connected knowledge base (in this case, Google Drive) for past proposals relevant to the entered topic.

• It retrieves the most relevant documents, such as previous airport expansion, transit system, highway, and bridge proposals, focusing on sustainability and infrastructure improvements.

3. AI-Powered Summarization

• An advanced AI model (OpenAI) reviews the retrieved proposals and generates a concise summary, highlighting key points and sustainability features from the most relevant documents.

4. Email Preparation

• Another AI model formats the summary into a clear, easy-to-read HTML email.

• The recipient’s email address is automatically included.

5. Automated Email Delivery

• The formatted summary is sent directly to the specified recipient’s email address using a secure Gmail connection.

6. Output Confirmation

• The workflow provides a confirmation that the email has been sent, along with a summary of the process.

Key Features:

• Automated research across your organization’s proposal archive.

• AI-generated summaries tailored to your new proposal topic.

• Seamless email delivery to any recipient you choose.

• No manual document review or formatting required.

Ideal for:
Teams preparing new proposals who want to leverage past work, ensure consistency, and save time on research and communication.