0
Loading
Back to Works
Case Study
2023

Mist

Moderation meets fun. A massive scale Discord bot serving 10M+ users.

Mist

Role

Co-Development

Client

Personal Project (Kars & Arex)

Stack

Next.jsDiscord.jsMongoDB

The Scale

Mist isn't just another bot; it's infrastructure. Built by Kars and external developer Arex, Mist was designed to handle scale from day one. Today, it serves over 10 million unique users across 1,400+ servers. Operating at this scale introduces challenges that most projects never face.

The Features

Mist bridges the gap between strict moderation tools and community engagement features. It provides server administrators with the tools they need to keep their communities safe—auto-moderation, logging, and role management—while also offering fun features that keep users engaged, such as leveling systems and economy games.

The Mist landing page

Technology

Handling 10 million users requires a database that can keep up. We chose MongoDB for its flexibility and performance with large, unstructured datasets typical of chat logs and user configurations.

Sharding & Performance

With Discord's gateway rate limits, a single process bot simply wouldn't work. We implemented a robust sharding system, distributing the bot's workload across multiple processes and even multiple servers. This ensured that if one shard went down or faced high load, the rest of the bot remained operational.

We optimized our database schemas to minimize read operations, heavily utilizing caching for user profiles and server settings. This reduced database load by over 60% during peak hours.

Note: The web interface for Mist is currently offline, but the bot itself remains fully operational and continues to serve millions of users.

Next Case Study

Loti

Start Project