Predictika · engineering software subscription quotes: product, licensing model, users, term, usage commit and add-on modules

An AI sales agent's engineering software subscription quote, checked before it goes out

An AI sales agent prepares an engineering software subscription quote for a fictional customer. Before it goes out, the agent sends it to Predictika’s validator, which checks it against the product, licensing model, user and term rules, compares the net total the agent states with the one it computes, and returns a verdict with every reason.

Scope: Bentley style engineering software subscription quotes: products, licensing models, users, term years, usage commit and add-on modules, with illustrative values in US dollars. See everything this validator covers

This validator models Bentley style engineering software subscription quotes: four products, three licensing models, the number of users, the term in years, a usage commit, add-on modules and discounts. It handles any quote made of these parts, not just the scenarios shown below.

  • Products and licensing models — MicroStation, OpenRoads Designer, OpenBuildings Designer and STAAD on a named user, Virtuoso usage or enterprise subscription, with STAAD not sold on the usage model.
  • Users and term — The number of users against the range of the model (up to 250 named, up to 2,000 usage, 250 to 20,000 enterprise) and the term against the longest term of the model.
  • Usage commit — A usage commit of at least USD 25 thousand, only on the usage model, with a finance review when it is under 20 percent of the estimated annual spend.
  • Add-on modules — Up to 6 modules on one quote, with a solution review above 4.
  • Discounts and order size — Volume and multi-year discounts plus any extra discount, approval by a sales director above 15 percent and a block above 25 percent, legal review of true-up terms above 2,000 enterprise users, and a deal desk review above USD 10,000,000.
  • The stated net total — A check that the net total the agent states matches the net total the validator computes, with an estimated price build-up in US dollars.

Not checked: other Bentley products and cloud services, real list prices, the actual true-up amounts of an enterprise agreement, partner pricing, tax, multiple currencies, and usage metering or invoicing after the sale. Built by Predictika from general public knowledge; prices, limits and percentages are illustrative estimates, not Bentley figures.

The AI sales agentBuilds the engineering software subscription 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.

7 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
Product and model pricesEasy
A price per product and a percentage per licensing model are simple catalog data.		The same data, checked together with every other rule.
Users and term limits per modelOnly with care
A different range for each licensing model means a validation rule per model or a lookup table.		The limits are read from the chosen model and compared on every call.
Products that are not sold on a modelOnly with workarounds
A product by model matrix has to be kept in option filters or rules and tends to drift.		The product is checked against the usage model flag of the chosen model.
Commit rules and the annual spendOnly with workarounds
A commit that must be at least a minimum, applies to one model only and is compared with an estimated annual spend needs several linked price and validation rules.		The commit is compared with the computed annual spend in the same pass as every other rule.
The total the agent statesOnly with workarounds
Comparing a stated total with a computed one means a separate rule per quote layout, and price rules rarely run on stated figures.		The stated net total is a field in the request and is checked against the computed total.
Every problem returned in one answerOnly with workarounds
Rules tend to fire as the quote is saved, 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 quoted net total must match the contract value the validator computes.
2The number of users must be within the minimum and maximum of the licensing model.
3Only some products are sold on the usage model.
4A term longer than the licensing model allows is blocked.
5More than 4 add-on modules need a solution review.

  

  

    
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. The total the agent states: Stated total below computed

    
The agent's request. MicroStation on the named user subscription for 25 users for 1 year with no add-on modules, Net total USD 63,000. Ready to send.

    
Verdict: Blocked.

    
    
      
  • Quoted Net Total. The quoted net total of USD 63000.0 does not match the USD 70000.0 the validator computed.
    • 
    

    
2. The total the agent states: Stated total corrected

    
The agent's request. MicroStation on the named user subscription for 25 users for 1 year with no add-on modules, Net total USD 70,000. Ready to send.

    
Verdict: Allowed.

    
3. Users: 300 users on a named subscription

    
The agent's request. MicroStation on the named user subscription for 300 users for 1 year with no add-on modules, Net total USD 798,000. Ready to send.

    
Verdict: Blocked.

    
    
      
  • User Limit. 300 users are above the 250 users of the Named user subscription.
    • 
    

    
4. Users: 250 users

    
The agent's request. MicroStation on the named user subscription for 250 users for 1 year with no add-on modules, Net total USD 665,000. Ready to send.

    
Verdict: Allowed.

    
5. Product and licensing model: STAAD on the usage model

    
The agent's request. STAAD on the Virtuoso usage subscription for 30 users for 1 year with a USD 100 thousand usage commit and no add-on modules, Net total USD 120,000. Ready to send.

    
Verdict: Blocked.

    
    
      
  • Usage Model Product. STAAD is not sold on the usage model.
    • 
    

    
6. Product and licensing model: STAAD on a named subscription

    
The agent's request. STAAD on the named user subscription for 30 users for 1 year with no add-on modules, Net total USD 96,000. Ready to send.

    
Verdict: Allowed.

    
7. Review by a person: Five add-on modules

    
The agent's request. MicroStation on the named user subscription for 50 users for 1 year with 5 add-on modules, Net total USD 245,000. Ready to send.

    
Verdict: Needs approval.

    
    
      
  • Module Review. 5 add-on modules are above 4; a solution review must confirm them.
    • 
    

  

  

    
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 an engineering software subscription quote for a fictional customer. Before it goes out, the agent sends it to Predictika’s validator, which checks it against the product, licensing model, user and term rules, compares the net total the agent states with the one it computes, and returns a verdict with every reason.
    2. 
      
  2. I will start with the first request by the agent.
    3. 
      
  3. The agent says. MicroStation on the named user subscription for 25 users for 1 year with no add-on modules, Net total USD 63,000. Ready to send.
    4. 
      
  4. Before it goes out, the agent calls the validator.
    5. 
      
  5. Blocked, with one finding, all in one answer. The quoted net total of USD 63000.0 does not match the USD 70000.0 the validator computed.
    6. 
      
  6. The agent states a net total of USD 63,000. From the same choices the validator computes USD 70,000 and returns one blocking finding for the difference of USD 7,000. The product, model, users, term and modules raised nothing; only the stated figure is wrong. Because the agent's figure travels in the request as the Claim_Net field, the validator can compare it with its own calculation instead of relying on the agent's account of its own quote.
    7. 
      
  7. What this shows: for the total the agent states, a rules-based CPQ can do this only with workarounds.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the stated net as to 70,000.
    10. 
      
  10. It calls the validator again.
    11. 
      
  11. This time the validator clears it.
    12. 
      
  12. The quote is unchanged. The agent now states USD 70,000, equal to the net total the validator computes, so the quote is allowed, and the line items reconcile to the total. The release call that follows was accepted: the validator answered "EXECUTED: quote released to the customer." (it is in the raw response).
    13. 
      
  13. One problem caught in one call, fixed in one pass, and a clean quote. That is the value of Predictika's validator: catching mistakes before they reach the customer. Now you can try the other scenarios yourself.
    14.