Predictika · Cloud database cluster orders
An AI sales agent prepares an order for a managed database cluster: a cluster tier, replica set or sharded cluster, backup, networking and security options, expected load and where it sits in the customer's project. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the documented tier and limit rules and returns a verdict, every reason and the figures it computed.
This validator models Atlas cluster configurations, not just the scenarios shown below.
Any Atlas cluster order that touches these rules is in scope — not only the scenarios pictured here. Not covered: prices and consumption rates (the validator produces no price build-up or line-item quote), tiers above M60, regions and cloud-provider availability, Atlas Search and other add-ons, Enterprise Advanced licenses, commit discounts and support plans.
7 scenarios in 3 groups · each group runs from the agent’s first attempt to the outcome · click one to switch
Illustrative agent statement. The validator receives the structured request below.
Proposing agent: AI sales agent (scripted)
Before sending, the agent must call Recorded run
What this means
Our assessment of what a typical rules-based CPQ (product rules, price rules and validation rules, as in Salesforce CPQ) does with the same cases. It is a judgment about effort and reliability, not a claim that nothing can be built.
| What the validator caught | Typical rules-based CPQ | Predictika |
|---|---|---|
| Features that exist only from a certain tier up | Only with care A product rule per feature and tier. It works, but every new tier or feature is another set of rules to keep in step with the documentation. | Tier, cluster type, backup, networking and security choices arrive in one request and every combination is checked together. |
| Limits that change with the tier | Only with care Data size, operations and connection limits differ by tier, so each limit needs its own table or rule and a comparison with the planned load. | Each limit is looked up from the selected tier on every call and compared with the planned load. |
| Node counts for replica sets and shards | Only with workarounds The data-bearing node total depends on the cluster type, shard count and nodes per shard, and a stated total is a free-entry field. | The total is computed from the shards and nodes on every call and compared with the figure the order states. |
| Limits across the whole project | Only with workarounds Rules usually see one order at a time; limits on how many clusters a project holds need the project count passed in and compared. | The project counts travel with the order and are checked together with the cluster settings. |
| Every problem returned in one answer | Only with workarounds Rules tend to fire as the order is saved or a step is completed, so a rep often fixes one error and meets the next. | All findings come back together, whatever order the choices arrived in. |
Every rule is checked on every call, in any order the choices arrive; one problem never hides another. Blocking findings stop the quote; approval findings need a person to sign off; everything else is computed.
| 1 | A sharded cluster needs cluster tier M30 or higher; Atlas does not offer sharding on the selected tier. |
| 2 | Backups cannot be enabled on a Free cluster. |
| 3 | Flex clusters do not support Continuous Backup or Point-in-Time Restore. |
| 4 | Free and Flex clusters do not support private endpoints or network peering; private networking needs M10 or higher. |
| 5 | Free and Flex clusters do not support encryption at rest with customer key management. |
| 6 | Free and Flex clusters do not support database auditing. |
| 7 | Multi-region and multi-cloud clusters are not available on Free or Flex clusters; they need M10 or higher. |
| 8 | Each replica set or shard needs 3, 5 or 7 electable nodes; the order asks for 4. |
| 9 | Each replica set or shard needs 3, 5 or 7 electable nodes; the order asks for 9. |
| 10 | A Free cluster is a 3-node replica set and the node count cannot be changed; the order asks for 5. |
| 11 | A project can have only one Free cluster; the project would have 2. |
| 12 | An Atlas project can hold at most 25 clusters; the project would have 26. |
| 13 | A Free cluster stores at most 0.5 GB of data; the order plans 600 MB. |
| 14 | A Flex cluster stores at most 5 GB of data; the order plans 6,144 MB. |
| 15 | A Free cluster is limited to 100 operations per second; the order plans 150. |
| 16 | A Flex cluster is limited to 500 operations per second (reads and writes combined); the order plans 600. |
| 17 | The order plans 2,000 simultaneous connections per node, but the selected tier allows 1,500. |
| 18 | The order plans 600 simultaneous connections per node, but the selected tier allows 500. |
| 19 | The order states 5 data-bearing nodes, but the cluster settings give 3. |
| 20 | The order states 3 data-bearing nodes, but the cluster settings give 12. |
| 21 | A Free cluster is a 3-node replica set and the node count cannot be changed; the order asks for 4. |
| 22 | A Free cluster stores at most 0.5 GB of data; the order plans 2,000 MB. |
| 23 | A Free cluster is limited to 100 operations per second; the order plans 900. |
| 24 | The order plans 900 simultaneous connections per node, but the selected tier allows 500. |
| 25 | The order states 6 data-bearing nodes, but the cluster settings give 8. |
The recorded result of the validator for each scenario on this page: the request the AI agent sent, the verdict, and every finding, in the order they appear above.
The agent's request. Please quote a replica set cluster on the Free tier with 3 nodes, Cloud Backup and a private endpoint, with customer key management. It will hold about 2,000 MB of data, serve 250 operations per second and open up to 400 connections per node, in a project that would have 3 clusters. The order states 3 data-bearing nodes in total. It is ready to send.
Verdict: Blocked.
The agent's request. Please quote a replica set cluster on the M10 tier with 3 nodes, Cloud Backup and a private endpoint, with customer key management. It will hold about 2,000 MB of data, serve 250 operations per second and open up to 400 connections per node, in a project that would have 3 clusters. The order states 3 data-bearing nodes in total. It is ready to send.
Verdict: Allowed.
The agent's request. Please quote a sharded cluster on the M20 tier with 4 shards of 4 nodes each, Cloud Backup and a private endpoint, with database auditing. It will hold about 1,500,000 MB of data, serve 15,000 operations per second and open up to 8,000 connections per node, in a project that would have 9 clusters. The order states 12 data-bearing nodes in total. It is ready to send.
Verdict: Blocked.
The agent's request. Please quote a sharded cluster on the M30 tier with 4 shards of 5 nodes each, Cloud Backup and a private endpoint, with database auditing. It will hold about 1,500,000 MB of data, serve 15,000 operations per second and open up to 8,000 connections per node, in a project that would have 9 clusters. The order states 12 data-bearing nodes in total. It is ready to send.
Verdict: Blocked.
The agent's request. Please quote a sharded cluster on the M40 tier with 4 shards of 3 nodes each, Cloud Backup and a private endpoint, with database auditing. It will hold about 1,500,000 MB of data, serve 15,000 operations per second and open up to 8,000 connections per node, in a project that would have 9 clusters. The order states 12 data-bearing nodes in total. It is ready to send.
Verdict: Allowed.
The agent's request. Please quote a replica set cluster on the Flex tier with 3 nodes, spread over 2 regions, Continuous Backup and VPC peering. It will hold about 4,000 MB of data, serve 800 operations per second and open up to 300 connections per node, in a project that would have 27 clusters. The order states 3 data-bearing nodes in total. It is ready to send.
Verdict: Blocked.
The agent's request. Please quote a replica set cluster on the M20 tier with 3 nodes, Continuous Backup and VPC peering. It will hold about 4,000 MB of data, serve 800 operations per second and open up to 1,500 connections per node, in a project that would have 24 clusters. The order states 3 data-bearing nodes in total. It is ready to send.
Verdict: Allowed.
The text of the narrated walkthrough, step by step.