Predictika · Cloud database quotes

An AI sales agent's Atlas cluster quote, checked before it goes out

An AI sales agent prepares a monthly quote for a team that wants MongoDB Atlas clusters: a Flex or Dedicated plan, the cluster tier, the storage, and the hours the clusters run. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the published Atlas pricing and limits and returns a verdict, every reason and the figures it computed.

Scope: MongoDB Atlas Flex and Dedicated cluster quotes (M20 to M700): published hourly rates, storage ranges per tier, cluster and hour limits, and the stated total. See everything this validator covers

This validator models monthly quotes for MongoDB Atlas Flex and Dedicated clusters as published on the MongoDB Atlas pricing page and Atlas documentation, not just the scenarios shown below.

  • Dedicated clusters — the monthly estimate is the published hourly list rate of the tier (M20 $0.20, M30 $0.54, M40 $1.04, M50 $2.00, M60 $3.95, M80 $7.30, M140 $10.99, M200 $14.59, M300 $21.85, M400 $22.40, M700 $33.26 an hour) times the hours billed and the number of clusters. Storage must lie in the range the pricing page lists for the tier, one replica set or shard holds at most 4 TB, and a project holds 1 to 25 clusters.
  • Flex clusters — priced by operations a second ($8, $15, $21, $26 or $30 a month for 0 to 100, 100 to 200, 200 to 300, 300 to 400 and 400 to 500 operations a second), with 5 GB of storage; more operations or storage needs a Dedicated cluster.
  • Approvals — Enterprise or Platinum support (priced through MongoDB sales), sharded clusters (not itemized on the public pricing page), and production workloads on Flex or M20 (MongoDB recommends M30 or larger for production) go to a person.

Any Atlas Flex or Dedicated quote that touches these rules is in scope — not only the scenarios pictured here. Not covered: the Free tier, the M10 tier (its exact hourly rate is not stated consistently on the pricing page), Enterprise Advanced licenses, support plan prices, backup, Atlas Search and Vector Search, Stream Processing, Data Federation, Online Archive, data transfer, Atlas Infinite, taxes, and negotiated or commit discounts.

The AI sales agentBuilds the Atlas cluster quote from the conversation and says it is ready.
Predictika validatorChecks every rule at once and returns the verdict, each finding, and the computed figures.

8 scenarios in 3 groups · each group runs from the agent’s first attempt to the outcome · click one to switch

What the agent proposes

The AI sales agent says

Illustrative agent statement. The validator receives the structured request below.

Proposing agent: AI sales agent (scripted)

Before sending, the agent must call Recorded run

Predictika validator


      

        
        
        
      

      
Waiting for the call.

      
    

  

  
  

    
Where rules-based CPQ gets it wrong
Click to expand
    
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 caughtTypical rules-based CPQPredictika
Storage that must fit the range of the chosen tierOnly with care
Each tier has its own minimum and maximum storage; a rule per tier works, but the table has eleven rows that must be kept in step with the pricing page.		The tier's range is looked up and compared with the storage on every call, and every out-of-range value is reported.
A monthly price built from an hourly rateOnly with care
A price rule multiplies the rate by hours and clusters; it works, but a stated total typed by a rep has to be compared with it separately.		The total is computed from the rate, hours and clusters and compared with the stated total in the same pass as the other rules.
A plan with its own limits next to another planOnly with workarounds
Flex has operations and storage limits that Dedicated does not, so the limits depend on the plan chosen, and each plan needs its own validation rules.		The plan is one of the choices, and only the limits of that plan are applied, together with all the others.
A storage ceiling that changes with shardingOnly with workarounds
More than 4 TB needs additional shards, but sharded pricing is not itemized, so the quote has to be blocked in one case and sent to a person in another.		Storage per shard, the shard count and the approval rule are in one request and are all reported together.
Recommendations that need a person to confirmOnly with care
A production workload on a small tier is not an error but should be reviewed; it needs an approval step that is easy to leave out.		The approval is raised whenever production is chosen on Flex or M20, and the quote cannot be executed until it is approved.
Every problem returned in one answerOnly with workarounds
Rules tend to fire as the quote 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.

  

  

    
What this validator catches: findings from its test cases
Click to expand
    
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.

    
1The quote has 19 GB of storage; the chosen tier offers 20 to 256 GB.
2The quote has 257 GB of storage; the chosen tier offers 20 to 256 GB.
3The quote has 39 GB of storage; the chosen tier offers 40 to 512 GB.
4The quote has 513 GB of storage; the chosen tier offers 40 to 512 GB.
5The quote has 4,097 GB on one replica set or shard; the published maximum is 4 TB (4,096 GB), and more needs additional shards.
6The quote has 500 GB of storage; the chosen tier offers 750 to 4,096 GB.
7The quote has 0 clusters; a project holds 1 to 25 clusters.
8The quote has 26 clusters; a project holds 1 to 25 clusters.
9The quote bills 745 hours in a month; a month has at most 744 hours.
10The quote bills 0 hours in a month; a month has at most 744 hours.
11The quote expects 501 operations a second on Flex; Flex is priced for 0 to 500, so a Dedicated cluster is needed.
12The quote has 6 GB of storage on Flex; Flex includes 5 GB.
13The quote states a monthly total of $787.40; the published rates give $788.40.
14The quote states a monthly total of $788.41; the published rates give $788.40.
15Enterprise or Platinum support is priced through MongoDB sales and is not published; a person should confirm it.
16MongoDB describes Flex for development and testing and M20 for low-traffic production only (M30 or larger is recommended for production); a person should confirm the sizing.
17The quote asks for 3 shards; sharded cluster pricing is not itemized on the public pricing page, so a person should confirm it.
18The quote has 5,000 GB on one replica set or shard; the published maximum is 4 TB (4,096 GB), and more needs additional shards.
19The quote asks for 2 shards; sharded cluster pricing is not itemized on the public pricing page, so a person should confirm it.
20The quote has 30 clusters; a project holds 1 to 25 clusters.

  

  

    
Scenarios and results
Click to expand
    
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.

    
1. Dedicated M20 and storage: Three Dedicated M20 clusters for production with 300 GB of storage each

    
The agent's request. Please quote 3 MongoDB Atlas Dedicated M20 clusters, with 300 GB of storage each, billed for 730 hours in the month. The workload is for production. The quote states a monthly total of $438.00. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Storage outside the tier range. The quote has 300 GB of storage; the chosen tier offers 20 to 256 GB.
    • 
      
  • Production workload on Flex or M20. MongoDB describes Flex for development and testing and M20 for low-traffic production only (M30 or larger is recommended for production); a person should confirm the sizing.
    • 
    

    
2. Dedicated M20 and storage: The agent resubmits after fixing the errors

    
The agent's request. Please quote 3 MongoDB Atlas Dedicated M30 clusters, with 300 GB of storage each, billed for 730 hours in the month. The workload is for production. The quote states a monthly total of $1,182.60. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Quote summary. 3 cluster(s), estimated monthly total of $1,182.60 at the published list rates.
    • 
    

    
3. Flex limits and the stated total: Two Flex clusters at 650 operations a second with 8 GB of storage and a total that is too low

    
The agent's request. Please quote 2 MongoDB Atlas Flex clusters, with 8 GB of storage each, handling about 650 operations a second. The workload is for production. The quote states a monthly total of $30.00. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Flex above 500 operations a second. The quote expects 650 operations a second on Flex; Flex is priced for 0 to 500, so a Dedicated cluster is needed.
    • 
      
  • Flex storage over 5 GB. The quote has 8 GB of storage on Flex; Flex includes 5 GB.
    • 
      
  • Stated total does not match. The quote states a monthly total of $30.00; the published rates give $60.00.
    • 
      
  • Production workload on Flex or M20. MongoDB describes Flex for development and testing and M20 for low-traffic production only (M30 or larger is recommended for production); a person should confirm the sizing.
    • 
    

    
4. Flex limits and the stated total: The agent resubmits after fixing the errors

    
The agent's request. Please quote 2 MongoDB Atlas Flex clusters, with 5 GB of storage each, handling about 450 operations a second. The workload is for production. The quote states a monthly total of $60.00. It is ready to send.

    
Verdict: Needs approval.

    
    
      
  • Production workload on Flex or M20. MongoDB describes Flex for development and testing and M20 for low-traffic production only (M30 or larger is recommended for production); a person should confirm the sizing.
    • 
      
  • Quote summary. 2 cluster(s), estimated monthly total of $60.00 at the published list rates.
    • 
    

    
5. Flex limits and the stated total: The agent resubmits after fixing the errors

    
The agent's request. Please quote 2 MongoDB Atlas Dedicated M30 clusters, with 100 GB of storage each, billed for 730 hours in the month. The workload is for production. The quote states a monthly total of $788.40. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Quote summary. 2 cluster(s), estimated monthly total of $788.40 at the published list rates.
    • 
    

    
6. Large Dedicated order and premium support: Four M50 clusters with 5,000 GB, 800 hours, premium support and a stale total

    
The agent's request. Please quote 4 MongoDB Atlas Dedicated M50 clusters, with 5,000 GB of storage each, billed for 800 hours in the month. The customer wants Enterprise or Platinum support. The quote states a monthly total of $5,840.00. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Hours outside a month. The quote bills 800 hours in a month; a month has at most 744 hours.
    • 
      
  • More than 4 TB per replica set or shard. The quote has 5,000 GB on one replica set or shard; the published maximum is 4 TB (4,096 GB), and more needs additional shards.
    • 
      
  • Stated total does not match. The quote states a monthly total of $5,840.00; the published rates give $6,400.00.
    • 
      
  • Enterprise or Platinum support pricing. Enterprise or Platinum support is priced through MongoDB sales and is not published; a person should confirm it.
    • 
    

    
7. Large Dedicated order and premium support: The agent resubmits after fixing the errors

    
The agent's request. Please quote 4 MongoDB Atlas Dedicated M50 clusters, with 4,000 GB of storage each, billed for 730 hours in the month. The customer wants Enterprise or Platinum support. The quote states a monthly total of $5,840.00. It is ready to send.

    
Verdict: Needs approval.

    
    
      
  • Enterprise or Platinum support pricing. Enterprise or Platinum support is priced through MongoDB sales and is not published; a person should confirm it.
    • 
      
  • Quote summary. 4 cluster(s), estimated monthly total of $5,840.00 at the published list rates.
    • 
    

    
8. Large Dedicated order and premium support: The agent resubmits after fixing the errors

    
The agent's request. Please quote 4 MongoDB Atlas Dedicated M50 clusters, with 4,000 GB of storage each, billed for 730 hours in the month. The quote states a monthly total of $5,840.00. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Quote summary. 4 cluster(s), estimated monthly total of $5,840.00 at the published list rates.
    • 
    

  

  

    
Narration transcript
Click to expand
    
The text of the narrated walkthrough, step by step.

    
    
      
  1. Welcome to a demo of Predictika's Validator. An AI sales agent prepares a monthly quote for a team that wants MongoDB Atlas clusters: a Flex or Dedicated plan, the cluster tier, the storage, and more. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the published Atlas pricing and limits and returns a verdict, every reason and the figures it computed.
    2. 
      
  2. I will start with the first request by the agent.
    3. 
      
  3. The agent says. Please quote 4 MongoDB Atlas Dedicated M50 clusters, with 5,000 GB of storage each, billed for 800 hours in the month. The customer wants Enterprise or Platinum support. The rest of the request is on screen. The quote states a monthly total of $5,840.00.
    4. 
      
  4. Before it goes out, the agent calls the validator.
    5. 
      
  5. Blocked, with three findings and one approval, all in one answer. The quote bills 800 hours in a month; a month has at most 744 hours. The quote has 5,000 GB on one replica set or shard; the published maximum is 4 TB (4,096 GB), and more needs additional shards. The quote states a monthly total of $5,840.00; the published rates give $6,400.00.
    6. 
      
  6. The validator returned three blocking findings and one approval in one answer. A check that reads one field at a time would pass most of this; it shows only when the choices in the quote are checked against each other and against the stated total.
    7. 
      
  7. What this shows: for a monthly price built from an hourly rate, a rules-based CPQ is error-prone for this.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the storage to 4,000 GB and the hours billed in the month to 730.
    10. 
      
  10. It calls the validator again.
    11. 
      
  11. Not blocked, but held for a person.
    12. 
      
  12. The validator also returns the line-item quote, on screen.
    13. 
      
  13. The three blocking findings are cleared; one approval remains for a person: Enterprise or Platinum support pricing.
    14. 
      
  14. The agent resubmits once more.
    15. 
      
  15. The revised request. It changes the enterprise or Platinum support to no.
    16. 
      
  16. It calls the validator once more.
    17. 
      
  17. This time the validator clears it.
    18. 
      
  18. The validator also returns the line-item quote, on screen.
    19. 
      
  19. The earlier finding is cleared, and the validator's total, $5,840.00, matches the quote.
    20. 
      
  20. Four problems caught in one call, fixed in two passes, and a clean $5,840.00 quote. That is the value of Predictika's validator: catching mistakes before they reach the customer. Now you can try the other scenarios yourself.
    21.