Back Issues/Search Home → Calendar → Archive → RSS → Subscribe → Services → Vendors → Blog → Current Issue → Popular → Newsletters (LinkedIn, Substack)

All issues › Volume 344, Issue 2 › IT Vendor News › MongoDB

How Observe.AI Used MongoDB Resharding to Scale Its AI Platform

MongoDB, Thursday, November 12th, 2026

Observe.AI worked with MongoDB Consulting to reshard a constrained Atlas cluster nearing its disk limit and expand capacity while staying online.

Observe.AI, an agentic AI platform for contact centers, uses MongoDB Atlas as its data layer across more than 100 clusters and over 50 TB of data.

One production-critical sharded cluster had two data shards holding about 13 TB of compressed data each against a roughly 14 TB disk limit, with ingest of around 50 GB per day.

Working with MongoDB Consulting, the team added shard capacity and used MongoDB resharding as an online way to redistribute large collections.

They judged the default balancer too slow for the volume, since chunk migrations and range deletions would prolong pressure on the original shards, and the goal was a healthier horizontal layout with minimal production disruption.

more →  ·  More from MongoDB →