The Missing Piece Your AI-Powered App Needs!

Kais Radwan ❄️ — Fri, 24 May 2024 13:57:04 +0000

Hey everybody, Kais here!

I've been working on an open-source project for a while and am excited to tell you about it.

Lately, I've been integrating LLM-powered features into applications and found the process complex and time-consuming. From dealing with unexpected LLM outputs to constant data validation struggles and managing LLM flow, history, and server-client interactions—it can be a real headache.

That’s why I built Scoopika, an open-source platform that helps you integrate reliable AI agents into your applications effortlessly. These agents can collaborate and access external tools, all without the hassle.

Key Features of Scoopika

Here are some powerful features Scoopika offers out of the box and totally for free:

Custom AI Agents: Build AI agents that can collaborate, use external tools & APIs, and access real-time data based on context.
LLM Output Validation: Ensure full type-safety with LLM output validation.
Real-Time Streaming: Stream agent responses in real-time with over 10 hooks that work on both server and client-side with no additional setup.
Client-Side Actions: Empower agents to perform actions in the user's browser in real-time for a next-level user experience.
Managed Sessions & History: Keep track of conversations and maintain session history easily.
Secure API Keys: Connect your LLMs with your API keys safely stored on your servers.

Getting started

Let’s see how to get started with building AI agents using Scoopika.

Create a Free Account: Go to the Scoopika platform and sign up for a free account.
Create Your Agent: Simply give your agent a name, description, avatar, and prompt. You can choose the LLM powering your agent from a variety of available providers and models.
Generate a Token: Go here to generate a token for your agent.

Run agent

First, install the Scoopika package:

npm install @scoopika/scoopika

Then, initialize Scoopika and run your agent in your code:

And that’s it! Now you have an agent running on your server with all the built-in features mentioned above.

Web integration

We provide guides and examples for specific frameworks here.

If you want to use agents on the client-side, you need to create an API that uses the Scoopika container to handle requests with built-in caching and can communicate with the Scoopika client library. Here’s a simple example of creating a full API with Express:

Now you’re good to go! You can run your agents, manage sessions, and more, all from the client-side thanks to the route we added.

First, install the Scoopika client library:

npm install @scoopika/client

Then, see the code to run an agent:

Further exploration

I highly recommend you check out the documentation to learn more about equipping agents with tools, running multi-agent boxes, and adding agents as tools to other agents.

Got questions?

Feel free to contact us at anytime ;)