top of page
Search

5 ways to use the GPT API with no code

Updated: Aug 9, 2023


GPT API

You’ve likely already been wowed by the power of end-user AI tools like OpenAI’s chatGPT, Google’s Bard, Anthropic’s Claude or Jasper. Their ability to emulate human thought and manipulate data is mind blowing.


Most forward thinking companies are encouraging their employees to use generative AI to revise writing and idea generation. Now, the main question companies are asking is: How can we incorporate AI into our core business functions?


You might assume that you need a team of engineers to integrate these tools via an API. However, AI-savvy operators can use free or low cost tools to integrate with one of these APIs.


This post focuses on how and why to use the GPT API, all without writing any code.



Practical ways to use AI to add customer value and reduce cost

Companies across industries can benefit from integrating AI into their products and workflows. Here are 5 ways teams can implement the GPT API to either add customer value or reduce operating costs.



1) Finance: Log purchase orders as line items in a spreadsheet or database


When receiving purchase orders from various businesses, the format of each purchase order will usually differ based on the company sending it. Purchase orders may arrive as plain text emails, PDFs, excel spreadsheets, word documents, etc., each formatted slightly differently. This presents a challenge to finance and fulfillment teams who need to log purchase orders and fulfill them quickly. Manually entering these orders into spreadsheets or databases is not only labor-intensive but also susceptible to human error.


GPT can interpret each order, extracting essential details like vendor, quantity, and payment terms. This data can then populate spreadsheets, databases, or software like QuickBooks. Taking a step further, the fulfillment process can also be automated, creating standard notifications for in-house or external fulfillment teams.


2) Strategy: Monitor competitor offerings, pricing and discounts

Competitor Discount Email

Strategic decision makers need to stay in tune with their competitor's offerings, prices, and periodic discounts. Thankfully, most companies periodically send updates with this information. The key is consolidating this information and relaying it to decision makers in real time.


By funneling competitor emails and funneling them to a specific inbox, GPT can analyze incoming emails and catalog essential data points in a spreadsheet, Slack notifications, or a periodic email report. Using this data ensures that a company's offerings are compelling to a consumer who is likely shopping for the best offer they can find.


3) Customer Experience: Automatically respond to customer reviews

Online reviews can significantly impact a brand. Prompt, personalized feedback is essential for reflecting a brand's dedication to customer contentment. But addressing each review manually is time consuming and repetitive.


The GPT API can automatically draft a tailored response for each review, comprehending the user's sentiment and adjusting replies based on previous interactions.


Once this is set up, customers will get personalized and swift responses. Companies who haven’t kept up with online reviews in the past can feed GPT all prior reviews and retroactively respond to them. This way, future prospective customers can see the company’s dedication to responding to every customer review.



4) Marketing: Create customized drip emails based on user behavior

A drip campaign uses a series of emails to nurture and engage customers. Creating drip campaigns can be time consuming, and making updates to multiple email versions is tedious.


Instead of creating multiple versions of emails, GPT can draft a unique email for each client, referencing their engagement history and prior communications, ensuring continuity without repetition.



5) Sales: AI Generated Sales assets


AI generated pitch deck

Sales teams often use assets such as slide decks or one pagers to support a sale. Sales organizations must choose between using pre-made generic assets and spending time on manually updating personalized materials. Not only does this take time away from core sales activities, one copy/paste oversight can result in sending a client a deck with a competitor's name in it.


AI generated marketing brochure

AI can generate pitch decks using available client data, offering a personalized experience with minimal effort from a sales representative. When a lead reaches a certain stage, a CRM can push an update to GPT, asking it to create sales assets based on notes or other data fields captured during the sales process.



How to use the GPT API with no code

This all sounds great, but how do companies actually implement this, and without the use of code?


Create a Zapier account

Zapier acts as a bridge between numerous applications, automating workflows and tasks. For instance, think of a scenario where you collect data from a web form, record it in a spreadsheet, and then ping your team on Slack. Zapier can make all of this happen automatically. Zapier has a free tier that should allow you to start tinkering, and you can upgrade as needed.



Create an OpenAI account and add billing details

You may already be familiar with chatGPT, but to access the GPT API, you'll need an OpenAI account. This allows you to tap into the core model (GPT) that powers chatGPT, which is the online interface most people are using now.


Add Billing info

The billing information is used by OpenAI to charge you a fraction of a penny every time your tool interacts with the GPT API. The cost is extremely low. For example, based on the time of publishing this blog, running a process that sends 600 messages per month (20 messages per day) would cost about $1.50 per month for this kind of usage of the GPT API.


It's worth noting that if your team subscribes to chatGPT's premium version, priced at $20 per user per month, creating tools using the GPT API should be a much more cost effective practice. Plus, using the API helps standardize how team members are using chatGPT and allows for additional control over the model.


Generate a GPT API Key

OpenAI API Keys

An API key is a unique identifier that Zapier will use to communicate with the GPT API. You’ll need to generate one API key that Zapier will use to identify who is “calling” the GPT API. Once you have the API key from OpenAI, you can add the key to Zapier.





Start building “zaps”

Zapier interface

Using Zapier’s visual interface, define a “trigger” to initiate a process and pull in source information. This trigger could be set off by events like receiving a new purchase order,


online review, or competitor email.


Next, craft a prompt for GPT. An example prompt could be:


Take the following email and create a CSV with key values for: Company name, discount amount, expiration date.


Since large language models are excellent at manipulating text, GPT can format this in any format we need it in: CSV, JSON, markup table, etc. Zapier also has built in tools to format the data as needed.




The final step is to push the information into a usable format. For example, the new information can be added as a row to a google sheet. If desired, a Slack alert can be configured to alert the team that a new competitor offering has been logged.


Competitor offer tracker



Take the next step

Using a model like GPT is just one way you can do more with less by harnessing the power of AI. Thanks to off the shelf solutions like Zapier, this is now something all employees can (and should) be doing. If you want to explore how you and your teams can be using the GPT API, set up time and let’s chat!


122 views0 comments

Recent Posts

See All

Commenti


bottom of page