Predictika · Self-service ATM configurations

An AI sales agent's ATM configuration, checked before it goes out

An AI sales agent prepares a configuration for a cash-recycling ATM with its cassettes, recycled denominations, bunch size, optional check or coin module, memory and storage. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the published datasheet and returns a verdict, every reason and the figures it computed.

Scope: SelfServ 62 cash-recycling ATM configurations: cassettes and note capacity, denominations, bunch size, check or coin module, memory and storage. See everything this validator covers

This validator models SelfServ 62 cash-recycling ATM configurations, not just the scenarios shown below.

  • Cassettes and capacity — the Scalable Recycler 3.0 scales from 3 to 6 cassettes (full and split); the note capacity is totalled from the datasheet figures (recycling cassette up to 3,000 notes, deposit cassette up to 3,300, split cassette up to 700 plus 1,300) and compared with the capacity the quote states.
  • Denominations and bunch size — up to 8 recycling denominations and a bunch of up to 300 notes.
  • Check or coin module — the Check Processing Module (a bunch of up to 400 items) and the SmartHopper Coin Module (up to 11 denominations) are alternatives, and their figures need the module on the quote.
  • Platform — memory of 8, 16, 32 or 64 GB and a 240 or 480 GB SSD.

Any SelfServ 62 configuration that touches these rules is in scope — not only the scenarios pictured here. Not covered: prices (no price list is public, so this validator produces no price build-up or line-item quote), the other SelfServ models, safes, consumer-interface and security options, and managed services. Assumption: the 3 to 6 cassette limit is read as covering recycling, deposit and split cassettes together. The source is a datasheet that mentions Windows 10, so current models may differ.

The AI sales agentBuilds the ATM configuration 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
A cassette count that must stay within 3 to 6Only with care
The count is the sum of three cassette types on different lines, so a rule must add them before comparing.		The cassettes travel in one request; the total is computed and compared on every call.
Check and coin modules that exclude each otherOnly with care
A product rule between two options; it works, but each new module adds another pair to maintain.		Both options are in one request, so the conflict is found whichever order the choices arrive in.
Figures that need the module they belong toOnly with workarounds
A check bunch or coin denominations on a quote with no module is a cross-line condition that interacts with the other module rules.		The module and its figures are in one request and checked together.
A stated capacity that must equal the cassettesOnly with workarounds
The total is derived from several lines and then compared with a figure typed elsewhere; such a check is easy to leave out.		The capacity is computed from the cassettes and compared with the stated figure on every call.
Every problem returned in one answerOnly with workarounds
Rules tend to fire as the configuration 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 configuration has 2 cassettes; the recycler scales from 3 to 6 cassettes (full and split).
2The configuration has 7 cassettes; the recycler scales from 3 to 6 cassettes (full and split).
39 denominations are set to recycle; the recycler handles up to 8.
4A bunch of 301 notes is above the maximum bunch size of 300 notes.
5The Check Processing Module and the SmartHopper Coin Module are alternatives; only one of them can be selected.
6A check bunch of 401 items is above the maximum of 400 items.
7100 check items are quoted without the Check Processing Module.
812 coin denominations are set; the SmartHopper Coin Module handles up to 11.
95 coin denominations are quoted without the SmartHopper Coin Module.
1012 GB of memory is not an available size; the options are 8, 16, 32 and 64 GB.
11128 GB of memory is not an available size; the options are 8, 16, 32 and 64 GB.
12A 500 GB SSD is not an available size; the options are 240 and 480 GB.
13The quote states a maximum capacity of 15,000 notes, but the selected cassettes hold up to 14,300 notes.
14A bunch of 400 notes is above the maximum bunch size of 300 notes.

  

  

    
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. Cassettes, denominations and bunch size: Seven cassettes, nine recycling denominations and a 350 note bunch

    
The agent's request. Please configure a SelfServ 62 cash-recycling ATM with 5 recycling cassettes, 1 deposit cassette and 1 split cassette, 9 denominations set to recycle and a bunch size of 350 notes. The computer is an Intel Core i5 with 16 GB of memory and a 240 GB SSD. The quote states that the cassettes hold up to 20,300 notes. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Cassette count outside 3 to 6. The configuration has 7 cassettes; the recycler scales from 3 to 6 cassettes (full and split).
    • 
      
  • Too many recycling denominations. 9 denominations are set to recycle; the recycler handles up to 8.
    • 
      
  • Bunch size too large. A bunch of 350 notes is above the maximum bunch size of 300 notes.
    • 
    

    
2. Cassettes, denominations and bunch size: The agent resubmits after fixing the errors

    
The agent's request. Please configure a SelfServ 62 cash-recycling ATM with 4 recycling cassettes, 1 deposit cassette and 1 split cassette, 8 denominations set to recycle and a bunch size of 300 notes. The computer is an Intel Core i5 with 16 GB of memory and a 240 GB SSD. The quote states that the cassettes hold up to 17,300 notes. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Configuration summary. 6 cassettes holding up to 17,300 notes; 8 denominations set to recycle.
    • 
    

    
3. Check and coin modules: Check and coin modules together with a 450 item check bunch

    
The agent's request. Please configure a SelfServ 62 cash-recycling ATM with 3 recycling cassettes, 1 deposit cassette and 1 split cassette, 6 denominations set to recycle and a bunch size of 250 notes, with the Check Processing Module handling a bunch of 450 items, and the SmartHopper Coin Module. The computer is an Intel Core i5 with 16 GB of memory and a 240 GB SSD. The quote states that the cassettes hold up to 14,300 notes. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Check and coin modules together. The Check Processing Module and the SmartHopper Coin Module are alternatives; only one of them can be selected.
    • 
      
  • Check bunch too large. A check bunch of 450 items is above the maximum of 400 items.
    • 
    

    
4. Check and coin modules: The agent resubmits after fixing the errors

    
The agent's request. Please configure a SelfServ 62 cash-recycling ATM with 3 recycling cassettes, 1 deposit cassette and 1 split cassette, 6 denominations set to recycle and a bunch size of 250 notes, with the SmartHopper Coin Module set up for 12 coin denominations, with a check bunch of 450 items. The computer is an Intel Core i5 with 16 GB of memory and a 240 GB SSD. The quote states that the cassettes hold up to 14,300 notes. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Check bunch too large. A check bunch of 450 items is above the maximum of 400 items.
    • 
      
  • Check items without the check module. 450 check items are quoted without the Check Processing Module.
    • 
      
  • Too many coin denominations. 12 coin denominations are set; the SmartHopper Coin Module handles up to 11.
    • 
    

    
5. Check and coin modules: The agent resubmits after fixing the errors

    
The agent's request. Please configure a SelfServ 62 cash-recycling ATM with 3 recycling cassettes, 1 deposit cassette and 1 split cassette, 6 denominations set to recycle and a bunch size of 250 notes, with the SmartHopper Coin Module set up for 11 coin denominations. The computer is an Intel Core i5 with 16 GB of memory and a 240 GB SSD. The quote states that the cassettes hold up to 14,300 notes. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Configuration summary. 5 cassettes holding up to 14,300 notes; 6 denominations set to recycle.
    • 
    

    
6. Capacity, memory and storage: A capacity that does not match the cassettes, with unavailable memory and SSD sizes

    
The agent's request. Please configure a SelfServ 62 cash-recycling ATM with 4 recycling cassettes, 1 deposit cassette and 1 split cassette, 8 denominations set to recycle and a bunch size of 300 notes. The computer is an Intel Core i7 with 12 GB of memory and a 500 GB SSD. The quote states that the cassettes hold up to 18,000 notes. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Memory is not an available size. 12 GB of memory is not an available size; the options are 8, 16, 32 and 64 GB.
    • 
      
  • Drive is not an available size. A 500 GB SSD is not an available size; the options are 240 and 480 GB.
    • 
      
  • Stated note capacity differs from the cassettes. The quote states a maximum capacity of 18,000 notes, but the selected cassettes hold up to 17,300 notes.
    • 
    

    
7. Capacity, memory and storage: The agent resubmits after fixing the errors

    
The agent's request. Please configure a SelfServ 62 cash-recycling ATM with 4 recycling cassettes, 1 deposit cassette and 1 split cassette, 8 denominations set to recycle and a bunch size of 300 notes. The computer is an Intel Core i7 with 16 GB of memory and a 480 GB SSD. The quote states that the cassettes hold up to 17,300 notes. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Configuration summary. 6 cassettes holding up to 17,300 notes; 8 denominations set to recycle.
    • 
    

  

  

    
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 configuration for a cash-recycling ATM with its cassettes, recycled denominations, bunch size, optional check or coin module, memory and storage. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the published datasheet 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 configure a SelfServ 62 cash-recycling ATM with 5 recycling cassettes, 1 deposit cassette and 1 split cassette, 9 denominations set to recycle and a bunch size of 350 notes. The computer is an Intel Core i5 with 16 GB of memory and a 240 GB SSD. The rest of the request is on screen. The quote states that the cassettes hold up to 20,300 notes.
    4. 
      
  4. Before it goes out, the agent calls the validator.
    5. 
      
  5. Blocked, with three findings, all in one answer. The configuration has 7 cassettes; the recycler scales from 3 to 6 cassettes (full and split). 9 denominations are set to recycle; the recycler handles up to 8. A bunch of 350 notes is above the maximum bunch size of 300 notes.
    6. 
      
  6. The validator returned three 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.
    7. 
      
  7. What this shows: for a cassette count that must stay within 3 to 6, a rules-based CPQ is error-prone for this.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the recycling cassettes to 4, the denominations set to recycle to 8, the bunch size to 300 notes and the stated maximum note capacity to 17,300 notes.
    10. 
      
  10. It calls the validator again.
    11. 
      
  11. This time the validator clears it.
    12. 
      
  12. All three earlier findings are cleared, and the validator's figure matches the quote.
    13. 
      
  13. Three problems 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.
    14.