Intro to Discuro

What can you do with Discuro?

Welcome to the Discuro API documentation!

 

With our product, developers can now easily build, test, and consume OpenAI models 10x faster. Our API allows seamless integration with your existing development workflow, making it simple to add the power of OpenAI to your projects.

 

Start exploring our documentation to see how our API can enhance your development process. Let's build something great together with Discuro! We understand the importance of data transformation in machine learning projects. That's why our API allows for flexible input/output (I/O) definitions, making it easy to work with various data formats and structures. This flexibility means that you can quickly and easily transform your data to fit the requirements of the OpenAI models you're using, without having to write custom code.

 

How does Discuro work?

The vision is simple, we want to be the ultimate all-in-one solution for developers looking to build, test & consume AI work-flows, fast. So, we built a product that eliminates all headaches when building on top of OpenAI, so that you can focus on building a fantastic product.

  1. Find & add your OpenAI API key into Discuro
    1. Find your key here → https://beta.openai.com/account/api-keys
    2. Add it into your Discuro settings here → http://build.discuro.com/settings
    3. Done!
  1. Create basic or complex AI flows using Prompt Templates & Orchestrators
  1. Execute your flows via our API
    1. along with your inputs
    2. and any meta-data you’d like us to store against that execution
  1. Retrieve your generated data-set upon completion
    1. and never have to interface with OpenAI directly ever again
 
 

How do I use Discuro?

We’ve tried to make Discuro’s learning curve as simple as possible, but where there is value, there is complexity to some extent.

 

Discuro’s foundation is built upon the idea of “Prompt Templates”.

 

These are templates you set up in our UI, which allow you to define simple I/O completions via the OpenAI API.

 

To create more complex flows, you’ll chain templates together in something we call an “Orchestrator”. Think of this like a more complex template, which “orchestrates” the chaining of templates together, consistently.

 

What you’ll learn from this help guide

  • What a Prompt Template is
    • How to create a Prompt Template
    • How to test a Prompt Template (in-app)
  • What an Orchestrator is
    • How to create an Orchestrator
    • How to test a Orchestrator (in-app)
  • Invoking/executing work-flows via API
    • Authorization (with API key)
    • Adding meta-data to your executions
    • Providing inputs to your executions
  • Troubleshooting
    • Common problems

Last updated on January 12, 2023