Integrating OpenAI’s API with Zapier for automating your content creation

Hey, it’s TekkemCore’s first blog post (woo!). I’ve noticed a few customers and others on the internet mention Zapier integrations with the GPT-4 API, so here’s an abstracted real-world example of what can be done with these two tools (as well as a few others) for AI-enabled automation.

Let’s check this out! 

Overview

First, here’s a diagram below of what we’re building. The use case is turning your “shower thoughts” or good ideas that randomly occur to you throughout the day into actionable marketing content for social media platforms like LinkedIn.

An overview of what we'll be building

Now, let’s review the tools we’ll be using:

  1. Smartphone – We need a way to capture our thoughts via audio with something like a voice memos app.
  2. Google Drive – This is where we’ll store the audio files with those thoughts. 
  3. Zapier – will orchestrate the end-to-end workflow.
  4. OpenAI’s APIs – for transcribing our audio to text (Whisper) and then rephrasing that text for wider appeal.
  5. Airtable – We’ll store the transcribed and modified text (i.e. our LinkedIn posts) here for later analysis and records.
  6. Buffer – This will schedule our LinkedIn posts for us. 
  7. LinkedIn – Our account for posting will be here. 

Costs

Overall, I estimated a monthly cost of around $75 USD to run this thing end-to-end. The majority of that cost was around the Zapier monthly plan. I assumed 5-10 voice notes of 2 minutes each per day, which means we’d have around 900-1,800 Zapier tasks, leading us to need the professional plan. The overall breakdown looks as follows:

  • Zapier: $73.00 (monthly professional plan for 2,000 tasks)
  • Whisper API: $1.80 – $3.60
  • GPT-4o Mini API: $0.012375 – $0.02475 (Note: This is now free with Zapier’s professional plan)
  • Airtable: $0 (free plan)
  • Buffer: $0 (free plan)

Let’s Build

I’ll assume you’ve got accounts in all the tools listed above as well as the API key for OpenAI. We’ll focus on the Zapier and OpenAI pieces, since that’s the newest integration that a lot of people haven’t had the chance to play around with yet.

Here’s how to set up Zapier to transcribe your audio file (credit to Corbin Brown for an excellent tutorial):

Once done, here’s how to use GPT-4 from within Zapier itself (again credit to Corbin Brown):

The End Result

This saves an estimated 50-100 hours each month (assuming 5-10 voice memos per day) from this workflow, which is a substantial amount of time savings for under $100 bucks.

Drop a message to us if you’d like us to build something similar for you!

← Back to Blog