About

The aim of this project is to create a multi-agentic system that fetches meeting details through the Fireflies API and anonymizes them. This anonymized content will then be customized according to the target social media platform.

Fireflies.ai

The meeting details are fetched through the Fireflies API. Following is the flow for fetching meeting details:

• The user is prompted to input a start date and time, and an end date and time, both in the format DD-MM-YYYY HH:MM.

• The details of meeting that occurred between the specified period are fetched and displayed to the user.

• The user is then asked to select the meeting they desire. Fireflies' summary for that meeting is then fetched along with the meeting transcript.

• The user is asked if they want the AI Agent to analyze the meeting transcript as well for a more detailed summary. However, this consumes a lot of tokens.

Fireflies' summary and optionally the full meeting transcript is then passed on to the Summarizer.

The AI Agents

The AI Agents have been built using the LangChain and LangGraph frameworks.

The Summarizer

The Summarizer sums up and organizes meeting details into data that can be used to generate content for social media. It uses Fireflies' meeting summary and optionally the full meeting transcript to generate detailed summaries.

The Anonymizer

The summary generated by the Summarizer is passed on to the Anonymizer. All confidential data and identifiable information is removed from the summary while making sure that the essence of the meeting is not lost.

The Writer

The anonymized summary is passed on to the Writer. It generates a blog post from the anonymized summary that can be published wherever the user desires. The prompt can be modified for whatever the target platform is.

Required Packages

The use of a python virtual environment is highly recommended.

python3 -m venv .venv
source ./.venv/bin/activate

For the script/notebook to run correctly, the following packages need to be installed:

pip install -U langchain_community langchain-groq langgraph

.env File

Create a .env file in the root directory.

The contents of the .env file should be as follows:

GROQ_API_KEY=""
FIREFLIES_API_KEY=""
WORDPRESS_USERNAME = ""
WORDPRESS_APPLICATION_PASSWORD = ""

API Keys

To obtain the Groq API Key, you need to log in to console.groq.com and then head to the 'API Keys' section.

To obtain the Fireflies API Key, you need to log in to Fireflies.ai and head to the settings page.

WordPress Rest API

Using the WordPress Rest API for publishing the Agent's blog requires the following:

• WordPress Username
• WordPress Application Password
• The WordPress Site's URL

Obtaining the WordPress Application Password:

• Login to your WordPress Dashboard.
• Head to the Profile Section and scroll to the bottom.
• You will see the option to add a new application password. On proceeding, you will be asked to assign a name to this new application password.
• Once you do that, you will be able to see the newly generated application password. Make sure to copy it as you won't be able to see it again due to security reasons.
• Copy it over to WORDPRESS_APPLICATION_PASSWORD in the .env file.

Please make sure that you do not use your login password for REST API Authentication.

You will also have to copy over your login username to WORDPRESS_USERNAME in the .env file.

Running the Script

cd src/
python3 main.py