Awesome
Mindwave: AI Chatbots, Agents & Document Q&A in Laravel Simplified.
WARNING: This package is NOT ready to be used yet!
Please follow @helgesverre for updates, and keep an eye on TODO.md to track progress.
Installation
You can install the package via composer:
composer require mindwave/mindwave
You can publish the config file with:
php artisan vendor:publish --tag="mindwave-config"
What is Mindwave?
Mindwave is a Laravel package that lets you easily build AI-powered chatbots, agents, and document question and answering (Q&A) functionality into your application.
With Mindwave, you can incorporate the power of OpenAI's state-of-the-art language models, Pinecone's vector search capabilities and your own custom "tools" to create intelligent software applications.
Code Example
<?php
use Illuminate\Support\Facades\File;
use Mindwave\Mindwave\Facades\DocumentLoader;
use Mindwave\Mindwave\Facades\Mindwave;
use Mindwave\Mindwave\Memory\ConversationBufferMemory;
$agent = Mindwave::agent(
memory: ConversationBufferMemory::fromMessages([])
);
Mindwave::brain()
->consume(
DocumentLoader::fromPdf(
data: File::get("uploads/important-document.pdf"),
meta: ["name" => "Important document"],
)
)
->consume(
DocumentLoader::fromUrl(
data: "https://mindwave.no/",
meta: ["name" => "Mindwave Documentation"],
)
)
->consume(
DocumentLoader::make("My name is Helge Sverre")
);
$agent->ask("List the top 3 most important things in 'important document'");
$agent->ask("What is mindwave?");
$agent->ask("What is my name?");
Use Cases
- 💬 Chatbots: Building AI-powered chatbots to provide support to customers.
- 🤖 Agents: Developing intelligent agents to automate tasks within an application.
- ❓ Document Q&A: Creating document question and answering (Q&A) systems to extract insights from text.
Documentation
Full documentation can be found here.
Demo Application
To see a real world example of how the Mindwave package can be used, checkout the
demo application here.
Changelog
Please see CHANGELOG for more information on what has changed recently.
Credits
- Helge Sverre
- Probots.io for the Pinecone PHP Package
- Tim Kleyersburg for the Weaviate PHP Package
- PGVector team for the PGVector PHP package
- Yethee for the Tiktoken PHP Package
- hkulekci for the Qdrant PHP Package
- LangChain for serving as the inspiration
- All Contributors
License
The MIT License (MIT). Please see License File for more information.