The story behind ProtectBot.

The story behind ProtectBot.

Table of contents

No heading

No headings in the article.

So, a little background into myself. I’m Maxim King, a node.js developer from London specialising on backend projects and API’s.

Protect, was written to tackle the increasing issue of malicious attacks. Prior to Protect’s release me and my colleagues were constantly getting targeted and attacked with malicious links which aimed to compromise users on the popular platform Discord. We were constantly looking for ways to tackle this, but other bots didn’t offer all the features we needed and would lack in certain areas.

Protect, was started up by myself and Brian 6 months ago with a mission to bring security to the masses.

The system was written using a module structure, we started off with only two modules the External Module and Ai Module. The External module uses OSS domain lists and API’s to pull in malicious links which gets added to a internal database. The Ai module was build using a Neural Network (brain.js) and was trained extensively to detect possible malicious links and behaviours in users messages to generate a score for each message sent, but without logging any content off the message sent.

This was only the start, as Protect started to attract attention from large organisations so did we. We expanded our internal database 10x larger, and grew upon our Ai network massively.

The main focus of Protect pointed towards our Ai modules, as more and more malicious links were popping up our internal databases wasn't enough to stop all the zero day links.

What grew from a project soon became an enterprise system, providing protection to large servers. Currently we’re protecting over 130,000 users and counting and we’re not stopping there. We have a lot planned for the future, and if you want to add your say feel free to send us suggestions in our discord server: discord.gg/U7jtfcQAVu