It would be great if we could organize the projects and not have them all at the same level of organization.
We could use this hierarchy to put rights, consult costs ...
A bit like the idea of Azure's management group or GCP's "Folder".

Create databases (DB) and collections via API thorough Terraform after the cluster has been created in Atlas. This would provide the ease of writing and running it in a single script before any data is loaded.

mLab had the option to create a single-node deployment, which was very nice for development environments where redundancy isn't important, but keeping costs down is.

Would be nice to see this in Atlas as well.

Have an Export feature WITHIN Atlas UI (Data Explorer) where user can perform (select collection and export as .csv/.json directly from the website without using Compass/data dump) option for collection integrated into (JS) MongoDB Atlas Cloud.

Add a button/toggle to Hide and Unhide an Index from the Data Explorer/Indexes display for a 4.4+ deployment.

Add support to send cluster metrics to Elastic. Elastic has recently been putting a lot of work into their APM and logging solution, this is now a viable alternative to NewRelic and Datadog and is gaining traction. It would be great if Atlas supported sending metrics to Elastic. Please see below for references

https://www.elastic.co/apm
https://www.elastic.co/siem

The use case for this is the same as NewRelic or Datadog. I wish to monitor my Atlas clusters from inside Elastic, create dashboards and set up alerting

An option to protect the "Pause" action on clusters with 2FA would help to reduce the chance of accidental pausing of production systems.

Hi.
We would like to have the possibility to administrate user rights also for the Billing area. In order to restrict all users seeing the invoices and to give this right only to particular people who are directly responsible for it. Please consider such feature for the next releases.

Provide Peering support for "Classic" (Hub and Spoke) VNet Peering topology, either directly or via a NAT

Alerts can be configured to send to specific Atlas Users, but not Atlas Teams, which diminishes the value of setting up a team of users.

We need a way to easily ingest logs into Splunk. It needs to be all logs - node and web portal activity. Ideally a Splunk connector would be provided but a stop-gap measure would be the ability to configure Atlas to write all logs to an Azure storage account.

Lack of native log ingestion ability may be an Atlas deal killer for us.

With the Atlas API, there is functionality to add labels to various clusters like EC2. It would be great if that was exposed through the Atlas UI as well.

When trying to implement Atlas infrastructure for Organization/Project users, the Teams functionality is useless unless we can add pending users to a Team. Right now, if a new user doesn't login for two weeks, we can't assign them to a Team until two weeks after the implementation was supposed to be configured.

Session timeout value in Atlas UI defaults at 12h. Allow it to be configurable per Project or Organization, with the minimum value being 15 minutes. All Atlas users within an organization that have been idle for the duration of the configured session timeout should be automatically logged out.

Upload good tutorials in YouTube about connecting mongodb atlas with available three options.

Make EBS Burst Balance Metrics available within the Atlas UI. Currently it is not easy to 100% understand how much burst credits are being used and whether they have actually been depleted.

You can infer it from the IOPS metrics going above baseline IOPS and then suddenly dropping down but would be better to see the balance itself to be 100% sure. This will save support requests to do so.

Enable connectivity in Azure using private endpoints