What is Proficient AI?

Proficient AI is a platform that enables developers to add conversational AI agents to their apps in minutes. Our interaction APIs and SDKs coupled with our powerful web dashboard are the simplest way teams can add and operate -powered agents for their end-user applications.

How it works

Here is a simple overview of the steps to get started:

  1. Go to the web dashboard and create an agent with the specified model
  2. Customize the agent to your application’s needs by editing its configuration
  3. Add the prebuilt UI component to your app to connect it to the agent (3-5 lines of code)

Done! The agent appears in a chat window with a persistent conversation history and your users are ready to interact with it, once you deploy.


For more advanced features, you need to use the API, which lets you connect your agents to all parts of your application.

Demo

Here is a short demo video (2 min, 28 sec) that shows how you can add an AI agent to your React application.

This demo is from 7 Apr 2023. We’ve built a lot features since then and will add a new demo soon.

What can I build with it?

Proficient agents are abstract virtual robots that can be used for various purposes. In your application, you can present your agents in many forms such as:

  • chatbots that know your business and respond to customer queries
  • virtual tutors that teach your users a variety of subjects
  • AI assistants that engage with your product’s users

Basically, all types of applications that require an abstract AI entity to interact with their users can use a Proficient AI agent.

Reach out to us via Discord or email if you’re not sure whether you can use a Proficient agent for your project. We’d be more than happy to help you.

Architecture

We can visualize the architecture of a typical application powered by Proficient AI with the following diagram.

System architecture visualization

Key takeaways

There are a few key things to note here:

Proficient APIs

All operations on the Proficient platform are divided into two categories each of which is represented by a standalone API.

1. Client API

The Client API is designed for client applications that are allowed to access only a subset of all available operations. For example, agents can’t be configured from a client application. On the other hand, an authenticated user sending a message to an agent can be implemented with the Client API.

Client API Reference

View all Client API endpoints and operations that you can carry out from your client applications.

2. Admin API

The Admin API is used by backend applications. With admin-level access, you can use this API to create, configure and delete agents, monitor interactions, update user data and more.

Admin API Reference

View all Admin API endpoints and operations that you can carry out from privileged environments.