Predictika · Glass wine bottle orders

An AI sales agent's wine bottle order, checked before it goes out

An AI sales agent prepares an order for empty glass wine bottles from the public Vidrala wine bottle catalogue: the model, the colour, the number of whole pallets, the closing line the bottles must fit and whether the wine is still or sparkling. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the catalogue's published figures and returns a verdict, every reason and the figures it computed.

Scope: Vidrala empty glass wine bottles, the 58 models of the public catalogue's “Other markets” wine range: colours offered, capacity, finish, bottles per pallet and carbonatable models. See everything this validator covers

This validator models orders for the 58 wine bottle models in the public Vidrala catalogue (Other markets range, read October 2026), not just the scenarios shown below.

  • Model and colour — each model is offered only in the colours the catalogue lists for it (Flint, Eurogreen / Moss, Dark Green, Emerald Green, Dead Leaf).
  • Capacity — the volume the customer fills must equal the capacity the catalogue gives for the model (187 ml up to 1.5 L).
  • Whole pallets — orders are in whole pallets, starting at one; the validator computes the bottles per pallet and the bottles in total and compares them with the total the quote states.
  • Closure finish — the finish the customer's closing line needs (CETIE, CETIE PT, CARRÉE or BVP 25 H) must match the finish the catalogue lists for the model; where the catalogue lists no finish for a model, a person confirms it.
  • Sparkling wine — sparkling wine needs a model the catalogue's own filter marks as carbonatable.

Any wine bottle order for these catalogue models is in scope — not only the scenarios pictured here. Not covered: prices (Vidrala lists none and quotes on request, so this validator produces no price build-up or line-item quote), decoration and other options, made-to-order moulds, Vidrala's other glass packaging (beer, spirits, oils, juices, food), models of other market ranges, and delivery. Assumptions: the volume the customer fills is treated as the nominal capacity the catalogue lists; the closing line is checked only by its finish name against the catalogue finish; a model that the carbonatable filter does not return is treated as not carbonatable; no maximum number of pallets is modeled because the catalogue gives none.

The AI sales agentBuilds the bottle order from the customer's request 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
A different colour list for every modelOnly with care
A model-by-colour table of 58 models and five colours, with a rule for every pair that is not offered; easy to let one slip through.		The model and colour are looked up together and a colour the model is not offered in is stopped.
A pallet size that differs by modelOnly with care
A bottles-per-pallet figure per model, kept in step with the catalogue, and a total that has to be recomputed.		The bottles per pallet come from the model and the total is computed on every call.
A stated bottle total that disagrees with the palletsOnly with workarounds
The total the agent states is a separate field and needs an extra rule that compares it with pallets times bottles per pallet.		The stated total is part of the request and is compared with the computed total on every call.
Sparkling wine that needs a carbonatable bottleOnly with workarounds
The use of the wine on one line and a property of the bottle on another need a cross-line rule that interacts with the others.		Use and model are in one request and the carbonatable requirement is checked with the rest.
A closing line that must match the bottle finishOnly with workarounds
The finish depends on the model and the closing line on another choice, and models with no listed finish need their own handling.		The finish the model has is compared with the finish the customer needs, and an unlisted finish goes to a person.
Every problem returned in one answerOnly 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.

  

  

    
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 order has 0 pallets; the Vidrala catalogue quote form starts at one full pallet and goes up in whole pallets.
2BD LUXURY 75 CL is not offered in Flint in the Vidrala catalogue.
3FLÛTE ESTANDAR 75 CL is not offered in Emerald Green in the Vidrala catalogue.
4The customer fills 500 ml per bottle; BD NOVA NATURA 75 CL holds 750 ml.
5The customer fills 750 ml per bottle; BD HARMONIE 50 CL holds 500 ml.
6The order is for sparkling wine; the Vidrala catalogue does not list BD NOVA NATURA 75 CL as carbonatable.
7The order is for sparkling wine; the Vidrala catalogue does not list FLÛTE ESTANDAR 75 CL as carbonatable.
8The customer closing line needs the CARRÉE finish; the catalogue lists a different finish for BD NOVA NATURA 75 CL.
9The customer closing line needs the CETIE finish; the catalogue lists a different finish for BG CLASSIQUE 75 CL.
10The customer closing line needs the CETIE finish; the catalogue lists a different finish for BD ELEGANCIA NATURA 75 CL.
11The catalogue lists no finish for BD ELITE 75 CL BVS, so a person should confirm that it fits the CETIE closure the customer named.
12The quote states 7,000 bottles in total; 5 pallets of 1,624 bottles each give 8,120.
13The customer fills 500 ml per bottle; BD LUXURY 75 CL holds 750 ml.
14The order is for sparkling wine; the Vidrala catalogue does not list BD LUXURY 75 CL as carbonatable.
15The customer closing line needs the CARRÉE finish; the catalogue lists a different finish for BD LUXURY 75 CL.
16The quote states 7,000 bottles in total; 5 pallets of 1,392 bottles each give 6,960.
17BD ELITE 75 CL BVS is not offered in Emerald Green in the Vidrala catalogue.

  

  

    
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. Still wine in 75 cl bottles: Twelve pallets of a 75 cl bottle in a colour it is not offered in, with a wrong bottle total

    
The agent's request. Please quote 12 pallets of BD LUXURY 75 CL wine bottles in Flint glass, for a customer filling 750 ml per bottle with still wine, on a closing line that needs the CETIE finish. The quote states 16,000 bottles in total. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Colour not offered for this model. BD LUXURY 75 CL is not offered in Flint in the Vidrala catalogue.
    • 
      
  • Stated bottle total does not match. The quote states 16,000 bottles in total; 12 pallets of 1,392 bottles each give 16,704.
    • 
    

    
2. Still wine in 75 cl bottles: The agent resubmits after fixing the errors

    
The agent's request. Please quote 12 pallets of BD LUXURY 75 CL wine bottles in Eurogreen / Moss glass, for a customer filling 750 ml per bottle with still wine, on a closing line that needs the CETIE finish. The quote states 16,704 bottles in total. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Order summary. 12 pallets of BD LUXURY 75 CL, 1,392 bottles per pallet, 16,704 bottles in total.
    • 
    

    
3. Sparkling wine: Sparkling wine order on a flute that is not carbonatable, with the wrong closing line

    
The agent's request. Please quote 10 pallets of FLÛTE ESTANDAR 75 CL wine bottles in Dark Green glass, for a customer filling 750 ml per bottle with sparkling wine, on a closing line that needs the CARRÉE finish. The quote states 13,920 bottles in total. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Sparkling wine needs a carbonatable bottle. The order is for sparkling wine; the Vidrala catalogue does not list FLÛTE ESTANDAR 75 CL as carbonatable.
    • 
      
  • Closure finish does not match the bottle. The customer closing line needs the CARRÉE finish; the catalogue lists a different finish for FLÛTE ESTANDAR 75 CL.
    • 
    

    
4. Sparkling wine: The agent resubmits after fixing the errors

    
The agent's request. Please quote 10 pallets of FLUTE HIGH 75 CL wine bottles in Dark Green glass, for a customer filling 750 ml per bottle with sparkling wine, on a closing line that needs the CARRÉE finish. The quote states 13,380 bottles in total. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Closure finish does not match the bottle. The customer closing line needs the CARRÉE finish; the catalogue lists a different finish for FLUTE HIGH 75 CL.
    • 
    

    
5. Sparkling wine: The agent resubmits after fixing the errors

    
The agent's request. Please quote 10 pallets of FLUTE HIGH 75 CL wine bottles in Dark Green glass, for a customer filling 750 ml per bottle with sparkling wine, on a closing line that needs the CETIE finish. The quote states 13,380 bottles in total. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Order summary. 10 pallets of FLUTE HIGH 75 CL, 1,338 bottles per pallet, 13,380 bottles in total.
    • 
    

    
6. Bottle size and finish: A 50 cl bottle chosen for a 750 ml fill

    
The agent's request. Please quote 6 pallets of BD HARMONIE 50 CL wine bottles in Flint glass, for a customer filling 750 ml per bottle with still wine, on a closing line that needs the CETIE finish. The quote states 11,916 bottles in total. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Bottle capacity does not match the fill volume. The customer fills 750 ml per bottle; BD HARMONIE 50 CL holds 500 ml.
    • 
    

    
7. Bottle size and finish: The agent resubmits after fixing the errors

    
The agent's request. Please quote 6 pallets of BD ELITE 75 CL BVS wine bottles in Flint glass, for a customer filling 750 ml per bottle with still wine, on a closing line that needs the CETIE finish. The quote states 8,352 bottles in total. It is ready to send.

    
Verdict: Needs approval.

    
    
      
  • Catalogue lists no finish for this model. The catalogue lists no finish for BD ELITE 75 CL BVS, so a person should confirm that it fits the CETIE closure the customer named.
    • 
      
  • Order summary. 6 pallets of BD ELITE 75 CL BVS, 1,392 bottles per pallet, 8,352 bottles in total.
    • 
    

    
8. Bottle size and finish: The agent resubmits after fixing the errors

    
The agent's request. Please quote 6 pallets of BD ELITE 75 CL wine bottles in Flint glass, for a customer filling 750 ml per bottle with still wine, on a closing line that needs the CETIE finish. The quote states 8,352 bottles in total. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Order summary. 6 pallets of BD ELITE 75 CL, 1,392 bottles per pallet, 8,352 bottles in total.
    • 
    

  

  

    
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 order for empty glass wine bottles from the public Vidrala wine bottle catalogue: the model, the colour, the number of whole pallets, and more. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the catalogue's published figures 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 10 pallets of FLÛTE ESTANDAR 75 CL wine bottles in Dark Green glass, for a customer filling 750 ml per bottle with sparkling wine. The quote states 13,920 bottles in total. 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. The order is for sparkling wine; the Vidrala catalogue does not list FLÛTE ESTANDAR 75 CL as carbonatable. The customer closing line needs the CARRÉE finish; the catalogue lists a different finish for FLÛTE ESTANDAR 75 CL.
    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.
    7. 
      
  7. What this shows: for sparkling wine that needs a carbonatable bottle, a rules-based CPQ can do this only with workarounds.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the model to FLUTE HIGH 75 CL and the stated bottles in total to 13,380.
    10. 
      
  10. It calls the validator again.
    11. 
      
  11. Still blocked, with one finding, all in one answer. The customer closing line needs the CARRÉE finish; the catalogue lists a different finish for FLUTE HIGH 75 CL.
    12. 
      
  12. One finding cleared. Still blocking: closure finish does not match the bottle.
    13. 
      
  13. The agent resubmits once more.
    14. 
      
  14. The revised request. It changes the closing line finish to CETIE.
    15. 
      
  15. It calls the validator once more.
    16. 
      
  16. This time the validator clears it.
    17. 
      
  17. The earlier finding is cleared, and the validator's figure matches the quote.
    18. 
      
  18. Two problems caught in one call, fixed in two passes, 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.
    19.