Predictika · Cloud Commit Quote Check

Before an AI agent sends a cloud commit quote, the quote is checked against the vendor’s commit and discount rules

Quillon Cloud, a fictional cloud and observability vendor, lets an AI sales agent put together annual commit quotes for its Compute, Trace and Logs products. Before the agent sends one, the call goes to Predictika’s Cloud Commit Quote Check. It checks the product dependencies, the minimum commit and the term, adds the volume discount and the term bonus into one ceiling and tests the requested discount against it, compares the commit with the expected spend, checks the included containers, and checks that the annual total the agent stated matches the computed one. Then it returns one verdict: allow, needs approval, or block.

Illustrative only. Quillon Cloud and its customers are fictional. Product names, prices, commit minimums and discount limits are invented for this demonstration.

Scope: one annual commit quote, judged against product, commit, term, discount and usage rulesSee everything this validator covers

This validator decides one thing: whether one annual commit quote may be sent, given what is in it. It is not a pricing system, and it does not know the customer.

  • Actions — send one annual commit quote for Quillon Compute and Trace hosts on the Standard or Premier plan, log intake, searchable logs and prepaid containers, with a term of 12 to 36 months.
  • Rules — products that depend on other products, minimum commit, term, volume discount plus term bonus, discount authority, expected spend against the commit, included containers and the stated annual total.

Within these rules the validator handles any combination of the listed choices, not just the seven scenarios on this page. It never sees the customer, the contract or the order system; those facts are taken as stated.

The AI agentProposes sending a quote, says why it is fine, then calls the validator before sending.
Predictika validatorReads the quote, looks up the prices, computes the total, applies the rules, then returns one verdict.

7 scenarios in 4 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
A discount the contract does not allowOnly with workarounds
A quoting tool can hold a volume discount table and a term bonus. Adding the two for this host count and term, and then checking the stated total against the result, takes a rule that reads them together.		Blocked in the same call: discount above what volume and term allow; stated total does not match.
Products that depend on other productsOnly with care
Dependencies between products live in the price book notes. A rule has to read the host counts and the log lines together.		Blocked in the same call: tracing on more hosts than compute; searchable logs with no log intake; expected spend above the commit.
A commit under the minimumOnly with care
Most tools stop at the first failing check. Returning every finding together takes a rule set designed to do that.		Blocked in the same call: commit below the minimum; expected spend above the commit.
Nothing blocked, but a person must sign offOnly with workarounds
Discount authority has levels. A rule has to separate what is blocked from what only needs a sign-off.		Blocked in the same call: discount above what a rep may give.
Two things a person must look atOnly with care
Included allotments and commit headroom sit in the contract terms. A rule has to compute them from the hosts and the spend.		Blocked in the same call: commit far above expected spend; containers above the included allotment.

  

  

    
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 action; approval findings need a person to sign off; everything else is allowed.

    
1The quote has no products or usage.
2350 Trace hosts is more than the 300 Compute hosts; Trace is sold only on hosts that also run Compute.
31500 million searchable log events are quoted with no log intake; events must be taken in before they can be searched.
4A $20,000.00 annual commit is below the $25,000 minimum commit.
5A 6 month term is outside the 12 to 36 months the validator accepts.
612 percent discount exceeds the 8 percent maximum for 600 hosts on a 12 month term.
7Stated annual total does not match: the agent stated $510,480.00; the validator computes $449,222.40.
88 percent discount is above the 6 percent a sales rep can give; approvable up to 8 percent.
9Expected annual spend of $219,000.00 is above the $150,000.00 commit; usage beyond the commit is billed at on-demand rates.
10The $80,000.00 commit is more than 120 percent of the $48,000.00 expected annual spend; unused commit is not refunded.
111000 containers are above the 800 included with the hosts; the extra 200 are billed on demand, about USD 438.0 a month (an estimate).

  

  

    
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. Discount stacking: A 12 percent discount on a 12-month term

    
The agent's request. I'm sending Fenwick Retail a 12-month commit of 450,000 on 600 Premier compute hosts and 400 tracing hosts, with 20,000 GB of log intake and 3000 million searchable events, at a 12 percent discount. The annual total is 510,480.

    
Verdict: Blocked.

    
    
      
  • Discount above what volume and term allow. 12 percent discount exceeds the 8 percent maximum for 600 hosts on a 12 month term.
    • 
      
  • Stated total does not match. Stated annual total does not match: the agent stated $510,480.00; the validator computes $449,222.40.
    • 
    

    
2. Discount stacking: The agent resubmits after fixing the errors

    
The agent's request. I'm sending the same quote on a 36-month commit of 480,000 with a 6 percent discount. The annual total is $479,851.20.

    
Verdict: Allowed.

    
    
      
  • Price basis. Annual net $479,851.20; list $510,480.00; monthly equivalent $39,987.60; commit $480,000.00; Compute hosts 600; maximum discount 12 percent; included containers 4800.
    • 
    

    
3. Product dependencies: Tracing and searchable logs without their prerequisites

    
The agent's request. I'm sending Harlow Biotech a 24-month commit of 150,000 on 300 Standard compute hosts and 350 tracing hosts, with 1500 million searchable log events. The annual total is 219,000.

    
Verdict: Blocked.

    
    
      
  • Tracing on more hosts than compute. 350 Trace hosts is more than the 300 Compute hosts; Trace is sold only on hosts that also run Compute.
    • 
      
  • Searchable logs with no log intake. 1500 million searchable log events are quoted with no log intake; events must be taken in before they can be searched.
    • 
      
  • Expected spend above the commit. Expected annual spend of $219,000.00 is above the $150,000.00 commit; usage beyond the commit is billed at on-demand rates.
    • 
    

    
4. Product dependencies: The agent resubmits after fixing the errors

    
The agent's request. I'm sending the same quote with tracing on 300 hosts, 10,000 GB of log intake and a commit of 220,000. The annual total is $214,200.00.

    
Verdict: Allowed.

    
    
      
  • Price basis. Annual net $214,200.00; list $214,200.00; monthly equivalent $17,850.00; commit $220,000.00; Compute hosts 300; maximum discount 0 percent; included containers 1200.
    • 
    

    
5. Commit size: A commit below the minimum

    
The agent's request. I'm sending Eastbrook Bank a 12-month commit of 20,000 on 450 Standard compute hosts and 20,000 GB of log intake. The annual total is 104,400.

    
Verdict: Blocked.

    
    
      
  • Commit below the minimum. A $20,000.00 annual commit is below the $25,000 minimum commit.
    • 
      
  • Expected spend above the commit. Expected annual spend of $104,400.00 is above the $20,000.00 commit; usage beyond the commit is billed at on-demand rates.
    • 
    

    
6. Commit size: The agent resubmits after fixing the errors

    
The agent's request. I'm sending the quote again with a commit of 110,000 and an 8 percent discount. The annual total is $96,048.00.

    
Verdict: Needs approval.

    
    
      
  • Discount above what a rep may give. 8 percent discount is above the 6 percent a sales rep can give; approvable up to 8 percent.
    • 
      
  • Price basis. Annual net $96,048.00; list $104,400.00; monthly equivalent $8,004.00; commit $110,000.00; Compute hosts 450; maximum discount 8 percent; included containers 1800.
    • 
    

    
7. Usage and commit: Containers over the allotment, a large commit

    
The agent's request. I'm sending Pinecrest Games a 24-month commit of 80,000 on 200 Standard compute hosts with 10,000 GB of log intake and 1000 containers. The annual total is 48,000.

    
Verdict: Needs approval.

    
    
      
  • Commit far above expected spend. The $80,000.00 commit is more than 120 percent of the $48,000.00 expected annual spend; unused commit is not refunded.
    • 
      
  • Containers above the included allotment. 1000 containers are above the 800 included with the hosts; the extra 200 are billed on demand, about USD 438.0 a month (an estimate).
    • 
      
  • Price basis. Annual net $48,000.00; list $48,000.00; monthly equivalent $4,000.00; commit $80,000.00; Compute hosts 200; maximum discount 0 percent; included containers 800.
    • 
    

  

  

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

    
    
      
  1. Welcome to a demo of Predictika's Validator. Quillon Cloud, a fictional cloud and observability vendor, lets an AI sales agent put together annual commit quotes for its Compute, Trace and Logs products. Before the agent sends one, the call goes to Predictika’s Cloud Commit Quote Check. It checks the product dependencies, the minimum commit and the term, adds the volume discount and the term bonus into one ceiling and tests the requested discount against it, compares the commit with the expected spend, checks the included containers, and checks that the annual total the agent stated matches the computed one. Then it returns one verdict: allow, needs approval, or block.
    2. 
      
  2. I will start with the first request by the agent.
    3. 
      
  3. The agent says. I'm sending Fenwick Retail a 12-month commit of 450,000 on 600 Premier compute hosts and 400 tracing hosts, with 20,000 GB of log intake and 3000 million searchable events. The annual total is 510,480. The rest of the request is on screen.
    4. 
      
  4. Before it goes out, the agent calls the validator.
    5. 
      
  5. Blocked, with two findings, all in one answer. 12 percent discount exceeds the 8 percent maximum for 600 hosts on a 12 month term. Stated annual total does not match: the agent stated $510,480.00; the validator computes $449,222.40.
    6. 
      
  6. The validator returned two blocking findings 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 discount the contract does not allow, a rules-based CPQ can do this only with workarounds.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the term to 36 months, the annual commit to $480,000, the discount requested to 6 percent and the stated annual total to $479,851.20.
    10. 
      
  10. It calls the validator again.
    11. 
      
  11. This time the validator clears it.
    12. 
      
  12. Both earlier findings are cleared.
    13. 
      
  13. Two problems caught in one call, fixed in one pass, and a clean $479,851.20 quote. That is the value of Predictika's validator: catching mistakes before they reach the customer.
    14.