How to build a Newsletter Writer Agent

This agent automates the end-to-end process of searching internal knowledge bases, extracting key information, and drafting a professional investor newsletter ready for distribution.

Challenge

Manually gathering, summarizing, and formatting up-to-date fund performance, product updates, and market insights for investor newsletters is time-consuming and error-prone.

Industry

Finance

Department

Finance

Customer Success

Sales

Integrations

OpenAI

Google Docs

Workflow Overview

1. User Inputs

  • What product and performance do you want to focus on?
    Node: in-0
    The user types keywords or topics (e.g., “Multi-Asset Growth Fund Q3 performance”) to guide the newsletter’s focus.

  • Market Insights URL
    Node: in-1
    The user pastes a link to a market insights article or report to be summarized in the newsletter.

2. Data Retrieval from OneDrive

  • Search OneDrive (OAuth2)
    Node: knowledgebase-0
    Uses the user’s input (from in-0) to search OneDrive for relevant product documents (e.g., brochures, factsheets).

  • Search OneDrive (OAuth2) 1
    Node: knowledgebase-1
    Also uses the user’s input (from in-0) to search OneDrive, but specifically targets performance documents (e.g., performance reports).

3. Content Generation with AI (LLM Nodes)

  • Product Update Section
    Node: llm-1
    Summarizes product updates and key fund features using information from the product documents (knowledgebase-0) and the user’s focus (in-0).

  • Performance Section
    Node: llm-3
    Summarizes the latest fund performance using the performance documents (knowledgebase-1) and the user’s focus (in-0).

  • Market Insights Section
    Node: llm-2
    Summarizes the key points from the market insights article or report provided by the user (in-1).

  • Newsletter Opening
    Node: llm-0
    Generates a brief, engaging opening for the newsletter, referencing the outputs of the Performance Section (llm-3), Product Update Section (llm-1), Market Insights Section (llm-2), and the user’s original question (in-0).

4. Newsletter Assembly

  • Newsletter Drafter
    Node: llm-5
    Composes the full newsletter by combining the outputs of:

    • Newsletter Opening (llm-0)

    • Market Insights (llm-2)

    • Performance Section (llm-3)

    • Product Update Section (llm-1)

    This node is also set up to send the newsletter to Google Docs (using the “Create Google Doc” tool).

5. Output

  • Newsletter Output
    Node: out-0
    Presents the final, formatted investor newsletter to the user.

How the Data Flows

  1. User provides focus keywords and a market insights URL.

  2. The workflow searches OneDrive for relevant product and performance documents.

  3. AI nodes summarize and generate each section of the newsletter using the retrieved documents and user input.

  4. The newsletter is assembled and optionally sent to Google Docs.

  5. The final newsletter is displayed to the user.

Summary Table

Node Name/Type

Purpose

User Input

Collects focus keywords and market insights URL

Data Retrieval

Finds relevant product and performance docs in OneDrive

AI Summarization

Summarizes product updates, market insights, performance, and creates an opening

Assembly

Drafts the full newsletter and sends to Google Docs

Output

Displays the final newsletter

In summary:
This workflow takes your focus and a market insights link, finds the right documents, uses AI to summarize and draft each section, assembles a professional investor newsletter, and outputs it for you—optionally saving it to Google Docs.

Get started

Let’s Build AI Agents, Together

Book a demo to see how AI agents can help your team process unstructured documents and perform complex analysis faster and more accurately.

Get started

Let’s Build AI Agents, Together

Book a demo to see how AI agents can help your team process unstructured documents and perform complex analysis faster and more accurately.

Get started

Let’s Build AI Agents, Together

Book a demo to see how AI agents can help your team process unstructured documents and perform complex analysis faster and more accurately.