Predictika · Capsule filling machines

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

An AI sales agent prepares a configuration for a pharmaceutical capsule filling machine: the model, the required output, the products dosed into each capsule and the options. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the published machine data and returns a verdict, every reason and the figures it computed.

Scope: ADAPTA 100 and ADAPTA 200 capsule fillers: output, products per capsule, dosing units and options. See everything this validator covers

This validator models configurations of the ADAPTA 100 and ADAPTA 200 capsule fillers from the data IMA publishes, not just the scenarios shown below.

  • Models and output — the ADAPTA 100 is listed at up to 100,000 capsules an hour with 12 capsules per cycle, the ADAPTA 200 at up to 200,000 capsules an hour with 24 capsules per cycle. A required output above the model's maximum is rejected, and so are stated figures that differ from the listed ones.
  • Products per capsule — each dosing unit doses one product (powder, pellets, tablets, minitablets or liquid). The ADAPTA 100 is listed for up to 5 products in the same capsule and the ADAPTA 200 for up to 3; the brochure offers up to 5 products on request, so 4 or 5 products on the ADAPTA 200 go to a person.
  • Options — the vacuum bowl (for pre-compacting fine powder), the minitablet presence camera and the liquid container mixer and heating each need a matching powder, minitablet or liquid dosing unit in the configuration.

Any ADAPTA 100 or ADAPTA 200 configuration that touches these rules is in scope — not only the scenarios pictured here. Not covered: prices (IMA publishes none, so this validator produces no price build-up or line-item quote), the ADAPTA 50 and other IMA machines, capsule size and capsule type choices, weighing and rejection options, and format parts.

The AI sales agentBuilds the capsule filler 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 required output above the machine maximumOnly with care
Each model has its own maximum, so each needs its own threshold rule, kept in step with the product data.		The model sets its own maximum and every requested output is compared with it on every call.
A product count above what the model is listed forOnly with workarounds
The limit differs by model (5 on one, 3 on the other) and a middle band goes to a person, which takes several rules and an approval path.		The limit and the approval band are rules of the model, checked together with everything else.
An option that needs a matching dosing unitOnly with care
A dependency between an option and a count of dosing units of one type, usually written as one rule per option.		Each option is checked against the dosing units of the configuration on every call.
A stated figure that disagrees with the machine dataOnly with workarounds
The figure the agent states is free text or a separate field, so it takes an extra rule that compares it with the listed value.		The stated figures are part of the request and are compared with the listed ones 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 quote requires 100,001 capsules an hour; the selected model reaches at most 100,000 capsules an hour.
2The quote requires 200,001 capsules an hour; the selected model reaches at most 200,000 capsules an hour.
3The quote requires 150,000 capsules an hour; the selected model reaches at most 100,000 capsules an hour.
4The configuration doses 6 products into the same capsule; the ADAPTA doses at most 5.
5The ADAPTA 200 is listed for up to 3 products in the same capsule; 4 products are configured, which the IMA brochure offers only on request, so a person should confirm it with IMA.
6The ADAPTA 200 is listed for up to 3 products in the same capsule; 5 products are configured, which the IMA brochure offers only on request, so a person should confirm it with IMA.
7The configuration has 0 dosing units; at least 1 product must be dosed into the capsule.
8The vacuum bowl for pre-compacting fine powder is selected but the configuration has no powder dosing unit.
9The minitablet presence camera is selected but the configuration has no minitablet dosing unit.
10The liquid container mixer and heating option is selected but the configuration has no liquid dosing unit.
11The quote states 24 capsules per cycle; the selected model fills 12 capsules per cycle.
12The quote states 12 capsules per cycle; the selected model fills 24 capsules per cycle.
13The quote states a maximum output of 200,000 capsules an hour; the selected model is listed at 100,000 capsules an hour.
14The quote states a maximum output of 100,000 capsules an hour; the selected model is listed at 200,000 capsules an hour.
15The quote requires 250,000 capsules an hour; the selected model reaches at most 200,000 capsules an hour.

  

  

    
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. Dosing units and options: An ADAPTA 200 with 6 products and options without matching dosing units

    
The agent's request. Please configure an ADAPTA 200 capsule filler running at 180,000 capsules an hour, dosing 6 products into each capsule with 1 pellet dosing unit, 2 tablet dosing units, 1 minitablet dosing unit and 2 liquid dosing units, with the vacuum bowl and the minitablet presence camera. The configuration states 24 capsules per cycle and a maximum output of 200,000 capsules an hour. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Too many products in one capsule. The configuration doses 6 products into the same capsule; the ADAPTA doses at most 5.
    • 
      
  • Vacuum bowl without a powder dosing unit. The vacuum bowl for pre-compacting fine powder is selected but the configuration has no powder dosing unit.
    • 
    

    
2. Dosing units and options: The agent resubmits after fixing the errors

    
The agent's request. Please configure an ADAPTA 200 capsule filler running at 180,000 capsules an hour, dosing 4 products into each capsule with 1 pellet dosing unit, 1 tablet dosing unit, 1 minitablet dosing unit and 1 liquid dosing unit, with the vacuum bowl and the minitablet presence camera. The configuration states 24 capsules per cycle and a maximum output of 200,000 capsules an hour. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Vacuum bowl without a powder dosing unit. The vacuum bowl for pre-compacting fine powder is selected but the configuration has no powder dosing unit.
    • 
      
  • Four or five products on the ADAPTA 200. The ADAPTA 200 is listed for up to 3 products in the same capsule; 4 products are configured, which the IMA brochure offers only on request, so a person should confirm it with IMA.
    • 
    

    
3. Dosing units and options: The agent resubmits after fixing the errors

    
The agent's request. Please configure an ADAPTA 200 capsule filler running at 180,000 capsules an hour, dosing 3 products into each capsule with 1 powder dosing unit, 1 minitablet dosing unit and 1 liquid dosing unit, with the vacuum bowl and the minitablet presence camera. The configuration states 24 capsules per cycle and a maximum output of 200,000 capsules an hour. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Configuration summary. 3 product(s) dosed into each capsule; the selected model fills 24 capsules per cycle at up to 200,000 capsules an hour.
    • 
    

    
4. Model and output: An ADAPTA 100 at 150,000 capsules an hour with a wrong stated cycle

    
The agent's request. Please configure an ADAPTA 100 capsule filler running at 150,000 capsules an hour, dosing 3 products into each capsule with 2 powder dosing units and 1 pellet dosing unit. The configuration states 24 capsules per cycle. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Required output above the machine maximum. The quote requires 150,000 capsules an hour; the selected model reaches at most 100,000 capsules an hour.
    • 
      
  • Stated capsules per cycle does not match. The quote states 24 capsules per cycle; the selected model fills 12 capsules per cycle.
    • 
    

    
5. Model and output: The agent resubmits after fixing the errors

    
The agent's request. Please configure an ADAPTA 200 capsule filler running at 150,000 capsules an hour, dosing 3 products into each capsule with 2 powder dosing units and 1 pellet dosing unit. The configuration states 24 capsules per cycle. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Configuration summary. 3 product(s) dosed into each capsule; the selected model fills 24 capsules per cycle at up to 200,000 capsules an hour.
    • 
    

    
6. Products per capsule: An ADAPTA 200 with 4 products and a liquid mixer

    
The agent's request. Please configure an ADAPTA 200 capsule filler running at 90,000 capsules an hour, dosing 4 products into each capsule with 1 powder dosing unit, 1 pellet dosing unit, 1 minitablet dosing unit and 1 liquid dosing unit, with the liquid container mixer and heating. The configuration states 24 capsules per cycle and a maximum output of 200,000 capsules an hour. It is ready to send.

    
Verdict: Needs approval.

    
    
      
  • Four or five products on the ADAPTA 200. The ADAPTA 200 is listed for up to 3 products in the same capsule; 4 products are configured, which the IMA brochure offers only on request, so a person should confirm it with IMA.
    • 
      
  • Configuration summary. 4 product(s) dosed into each capsule; the selected model fills 24 capsules per cycle at up to 200,000 capsules an hour.
    • 
    

    
7. Products per capsule: The agent resubmits after fixing the errors

    
The agent's request. Please configure an ADAPTA 100 capsule filler running at 90,000 capsules an hour, dosing 4 products into each capsule with 1 powder dosing unit, 1 pellet dosing unit, 1 minitablet dosing unit and 1 liquid dosing unit, with the liquid container mixer and heating. The configuration states 12 capsules per cycle and a maximum output of 100,000 capsules an hour. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Configuration summary. 4 product(s) dosed into each capsule; the selected model fills 12 capsules per cycle at up to 100,000 capsules an hour.
    • 
    

  

  

    
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 pharmaceutical capsule filling machine: the model, the required output, the products dosed into each capsule and the options. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the published machine data 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 an ADAPTA 200 capsule filler running at 180,000 capsules an hour, dosing 6 products into each capsule with 1 pellet dosing unit. The configuration states 24 capsules per cycle and a maximum output of 200,000 capsules an hour. 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 configuration doses 6 products into the same capsule; the ADAPTA doses at most 5. The vacuum bowl for pre-compacting fine powder is selected but the configuration has no powder dosing unit.
    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 an option that needs a matching dosing unit, a rules-based CPQ is error-prone for this.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the dosing units, tablet to 1 and the dosing units, liquid to 1.
    10. 
      
  10. It calls the validator again.
    11. 
      
  11. Still blocked, with one finding and one approval, all in one answer. The vacuum bowl for pre-compacting fine powder is selected but the configuration has no powder dosing unit.
    12. 
      
  12. One finding cleared; still blocked on vacuum bowl without a powder dosing unit.
    13. 
      
  13. The agent resubmits once more.
    14. 
      
  14. The revised request. It sets the dosing units, powder to 1. It removes the dosing units, pellet and dosing units, tablet.
    15. 
      
  15. It calls the validator once more.
    16. 
      
  16. This time the validator clears it.
    17. 
      
  17. Both earlier findings are 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.
    19.