Predictika · Personal care dispensing systems

An AI sales agent's dispensing-system order, checked before it goes out

An AI sales agent prepares an order for a pump, sprayer, dispenser or foamer from a public product catalog, with a closure finish and the output per actuation. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the catalog's published specifications and returns a verdict, every reason and the figures it computed.

Scope: Silgan Dispensing personal care pumps, sprayers, dispensers and foamers (27 catalog products): closure finish, output per actuation and foam dose. See everything this validator covers

This validator models orders for 27 products in the public Silgan Dispensing Personal Care catalog (March 2026), not just the scenarios shown below.

  • Sprayers — SP22 (Panache Pulse), SP32, Sinfonia (Elegance), Sprayette IV, Mark IV, Mark VI, Mark VII Max, Mark VII Max Style, Mini-Trigger Curve and Romantica, each with the closure sizes and output per actuation the catalog lists.
  • Dispensers — D1, Harmony, P2000, ReVive 2cc, SD20C, SD200, SD400 and 800, and ShowerGuard, with their closure sizes and output (2 ml for most; 1.5 or 2 ml for the SD20C; 7 to 30 ml for the D1; 2 to 8 ml for the SD400 and 800; 4 ml for the ShowerGuard).
  • Foamers — EZ'R (fill volume by market), F2, F3, G3, M3, WRD4, WRM3, WRT4 and WRT6, with the closure size, the output code (L11, L9, L7 for full-size foamers; S6, S10 for travel size) and the foam dose each code gives.
  • Special cases — the inverted (upside-down) spray exists only on the Mark VII Max and gives 0.18 ml; some closure sizes are ribbed only, and the Mini-Trigger Curve closures are smooth wall only.

Any order for these products that touches these rules is in scope — not only the scenarios pictured here. Not covered: prices and minimum order quantities (the catalog lists none, so this validator produces no price build-up or line-item quote), the Healthcare, Beauty, Fragrance, Food and Home & Garden catalogs, deodorant packaging, aerosol actuators, caps and closures other than the closure finish, jars, decoration and colors, the Europa 5 and 15-415 finishes (the catalog text is ambiguous about them), and applications (the catalog lists them as suggestions).

The AI sales agentBuilds the dispensing-system order 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 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 closure finish the product is not offered withOnly with care
Each of the 27 products has its own list of closure sizes, so each needs its own compatibility table, and every catalog update means editing them by hand.		The product and the closure are in one request and compared with the catalog on every call.
An output per actuation the product does not offerOnly with care
Some products list a few discrete outputs, others a range, and the Mark VII Max has a separate inverted output, so the check is a mix of tables and ranges that must be kept in step.		Each product's outputs, ranges and the inverted exception are checked together on every call.
A foam dose that disagrees with the output codeOnly with workarounds
The foam dose the agent states is free text or a separate field, so it takes an extra rule that looks the dose up from the output code and compares.		The stated foam dose is part of the request and is compared with the dose the code gives on every call.
Ribbed-only closures, smooth-only closures and the inverted optionOnly with care
Each is a rule tied to one product or one closure size; they are easy to forget when a product is added.		All three are checked on every call, whatever order the choices arrived in.
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 selected closure finish is not among the closure sizes the catalog lists for the selected product.
2The order asks for 0.08 ml per actuation, which the catalog does not list for the selected product.
3The order asks for 0.15 ml per actuation, which the catalog does not list for the selected product.
4The order asks for 0.18 ml per actuation, which the catalog does not list for the selected product.
5The order asks for 0.21 ml per actuation, which the catalog does not list for the selected product.
6An inverted (upside-down) spray is listed in the catalog only for the Mark VII Max, not for the selected product.
7The catalog lists this closure size as ribbed only for the selected product, and the order specifies a smooth skirt.
8The order asks for 0.31 ml per actuation, which the catalog does not list for the selected product.
9The catalog lists the Mini-Trigger Curve closures as smooth wall only, and the order specifies a ribbed skirt.
10The order asks for 0.14 ml per actuation, which the catalog does not list for the selected product.
11The order asks for 31.0 ml per actuation, which the catalog does not list for the selected product.
12The order asks for 1.0 ml per actuation, which the catalog does not list for the selected product.
13The order asks for 2.0 ml per actuation, which the catalog does not list for the selected product.
14The EZR foamer is listed for fill volumes of 100 or 160 ml in the EU and 160 or 200 ml in the US; the order asks for 200 ml.
15The EZR foamer is listed for fill volumes of 100 or 160 ml in the EU and 160 or 200 ml in the US; the order asks for 100 ml.
16The order states 9.0 ml of foam per actuation, but the catalog lists 12.0 ml for the selected output code.
17The foam output code on the order is not one the catalog lists for the selected foamer (full-size foamers offer L11, L9 and L7; travel-size M3, WRD4 and WRM3 offer S6 and S10).
18The order states 9.0 ml of foam per actuation, but the catalog lists 4.4 ml for the selected output code.
19The order asks for 0.2 ml per actuation, which the catalog does not list for the selected product.

  

  

    
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. Fine mist sprayer: A Sinfonia sprayer with a 28-400 screw closure at 0.15 ml per actuation

    
The agent's request. Please order the Sinfonia sprayer with a 28-400 closure, at 0.15 ml per actuation. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Closure finish not offered for this product. The selected closure finish is not among the closure sizes the catalog lists for the selected product.
    • 
      
  • Output per actuation not offered. The order asks for 0.15 ml per actuation, which the catalog does not list for the selected product.
    • 
    

    
2. Fine mist sprayer: The agent resubmits after fixing the errors

    
The agent's request. Please order the Sinfonia sprayer with a 24-410 closure, at 0.13 ml per actuation. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Order summary. 0.13 ml per actuation ordered.
    • 
    

    
3. Upside-down spray: A Mark VI sprayer with the inverted (upside-down) option

    
The agent's request. Please order the Mark VI fine mist sprayer with a 22-415 closure with the inverted (upside-down) option, at 0.16 ml per actuation. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Inverted spray not offered. An inverted (upside-down) spray is listed in the catalog only for the Mark VII Max, not for the selected product.
    • 
    

    
4. Upside-down spray: The agent resubmits after fixing the errors

    
The agent's request. Please order the Mark VII Max fine mist sprayer with a 22-415 closure with the inverted (upside-down) option, at 0.18 ml per actuation. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Order summary. 0.18 ml per actuation ordered.
    • 
    

    
5. Foamer: A WRT4 foamer with a 30 mm closure, output code S6 and a stated 9 ml of foam

    
The agent's request. Please order the WRT4 foamer with a 30 mm screw-on closure and output code S6. The order states 9 ml of foam per actuation. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Closure finish not offered for this product. The selected closure finish is not among the closure sizes the catalog lists for the selected product.
    • 
      
  • Foam output code not offered. The foam output code on the order is not one the catalog lists for the selected foamer (full-size foamers offer L11, L9 and L7; travel-size M3, WRD4 and WRM3 offer S6 and S10).
    • 
      
  • Stated foam dose differs from the catalog. The order states 9.0 ml of foam per actuation, but the catalog lists 4.4 ml for the selected output code.
    • 
    

    
6. Foamer: The agent resubmits after fixing the errors

    
The agent's request. Please order the WRT4 foamer with a 40 mm screw-on closure and output code L11. The order states 9 ml of foam per actuation. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Order summary. Foamer with liquid dose 0.75 ml; the catalog lists 9.0 ml of foam per actuation for this output code.
    • 
    

    
7. Lotion pump: An SD200 dispenser with a 24-400 closure, a smooth skirt and 2.5 ml per actuation

    
The agent's request. Please order the SD200 dispenser with a 24-400 closure, at 2.5 ml per actuation. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Output per actuation not offered. The order asks for 2.5 ml per actuation, which the catalog does not list for the selected product.
    • 
      
  • Ribbed skirt required. The catalog lists this closure size as ribbed only for the selected product, and the order specifies a smooth skirt.
    • 
    

    
8. Lotion pump: The agent resubmits after fixing the errors

    
The agent's request. Please order the SD200 dispenser with a 24-400 closure and a ribbed skirt, at 2 ml per actuation. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Order summary. 2.0 ml per actuation ordered.
    • 
    

  

  

    
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 a pump, sprayer, dispenser or foamer from a public product catalog, with a closure finish and the output per actuation. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the catalog's published specifications 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 order the WRT4 foamer with a 30 mm screw-on closure and output code S6. The order states 9 ml of foam per actuation. The rest of the request is on screen.
    4. 
      
  4. Before it goes out, the agent calls the validator.
    5. 
      
  5. Blocked, with three findings, all in one answer. The selected closure finish is not among the closure sizes the catalog lists for the selected product. The foam output code on the order is not one the catalog lists for the selected foamer (full-size foamers offer L11, L9 and L7; travel-size M3, WRD4 and WRM3 offer S6 and S10). The order states 9.0 ml of foam per actuation, but the catalog lists 4.4 ml for the selected output code.
    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 closure finish the product is not offered with, a rules-based CPQ is error-prone for this.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the closure finish to 40 mm screw-on and the foam output code to L11.
    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. Now you can try the other scenarios yourself.
    14.