Predictika · Air filter orders

An AI sales agent's air filter order, checked before it goes out

An AI sales agent prepares an order for a Hi-Flo extended surface pocket filter: the efficiency, the size, the number of pockets, the design airflow, the system type and the figures it states on the quote. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the published product sheet and returns a verdict, every reason and the listed values it compared against.

Scope: Hi-Flo multi-pocket filters in MERV 14, 13, 11 and 9: the 40 listed size and pocket combinations, airflow range, stated media area, rated airflow and resistance, and the final resistance and temperature limits. See everything this validator covers

This validator models the selection chart and data notes published in the Camfil Hi-Flo product sheet, not just the scenarios shown below.

  • Listed models — the height, width, depth and pocket count must be a combination on the selection chart; anything else is blocked.
  • Airflow — the design airflow is compared with the listed range for the model, taken at 100 percent for constant volume systems and 80 percent for VAV systems; above the range is blocked and below it goes to a person.
  • Stated figures — a media area or rated airflow that differs from the chart is blocked, and an initial resistance lower than the listed value goes to a person for approval.
  • Limits from the data notes — a planned final resistance above 1.5 inches w.g., an operating temperature above 158 F, or a stated pressure capability above 5.0 inches w.g. is blocked; a final resistance above the recommended 1.0 inch goes to a person.
  • Handling notes — 32 and 36 inch deep filters without pocket loops, and the 20 by 24 header size (available only on consultation with the factory), go to a person.

Any order for these filters that touches these rules is in scope — not only the scenarios pictured here. Not covered: prices (Camfil publishes no price list for Hi-Flo, so this validator produces no price build-up or line-item quote), the other Hi-Flo variants and Camfil product lines, and the resistance at airflows other than the rated one.

The AI sales agentBuilds the filter 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 size and pocket chart that is not a formulaOnly with care
Only 40 height, width, depth and pocket combinations exist, each with its own airflow range and media area, so the product rules have to carry the whole table.		The chart is a lookup computed on every call, and an unlisted combination is found whichever order the choices arrive in.
Airflow limits that depend on the system typeOnly with workarounds
The same filter is checked at 100 percent of the design airflow for constant volume and 80 percent for VAV, so the comparison changes with a choice made on another line.		The system type sets the factor, and the listed range for the model is applied in the same pass.
Stated figures checked against the chartOnly with care
A quoted media area, rated airflow or resistance must match the chart entry for that exact model and efficiency, so each claim needs its own formula.		The listed values are computed on every call and compared with what the quote states.
Limits that need a person to confirmOnly with care
A final resistance between 1.0 and 1.5 inches, a missing pocket loop or a factory-only header size are not errors, so they need an approval step apart from the blocks.		Blocks and approvals are separate findings in one answer, and the Execute step is refused until each one is cleared.
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.

    
1A filter measuring 24 x 24 x 22 inches with 11 pockets is not a model on the Hi-Flo selection chart.
2A filter measuring 24 x 24 x 20 inches with 10 pockets is not a model on the Hi-Flo selection chart.
3A filter measuring 24 x 16 x 22 inches with 10 pockets is not a model on the Hi-Flo selection chart.
4A filter measuring 18 x 24 x 22 inches with 10 pockets is not a model on the Hi-Flo selection chart.
5A filter measuring 24 x 12 x 22 inches with 6 pockets is not a model on the Hi-Flo selection chart.
6The design airflow is 2,001 CFM, taken at 100 percent for this system type; the highest listed airflow for this model is 2,000 CFM.
7The design airflow is 999 CFM, taken at 100 percent for this system type; the lowest listed airflow for this model is 1,000 CFM, so a person should confirm the selection.
8The design airflow is 2,501 CFM, taken at 80 percent for this system type; the highest listed airflow for this model is 2,000 CFM.
9The design airflow is 1,874 CFM, taken at 80 percent for this system type; the lowest listed airflow for this model is 1,500 CFM, so a person should confirm the selection.
10The design airflow is 1,499 CFM, taken at 100 percent for this system type; the lowest listed airflow for this model is 1,500 CFM, so a person should confirm the selection.
11The quote states 80 square feet of media; the selection chart lists 73 square feet for this model.
12The quote states a rated airflow of 2,000 CFM; the selection chart lists 1,750 CFM as the rated (medium) airflow for this model.
13The quote states an initial resistance of 0.3 inches w.g.; the selection chart lists 0.36 inches w.g. at the rated airflow for this model and efficiency, so a person should confirm it.
14The quote states an initial resistance of 0.5 inches w.g.; the selection chart lists 0.54 inches w.g. at the rated airflow for this model and efficiency, so a person should confirm it.
15The quote states an initial resistance of 0.2 inches w.g.; the selection chart lists 0.22 inches w.g. at the rated airflow for this model and efficiency, so a person should confirm it.
16The planned final resistance is 1.2 inches w.g.; the recommended final resistance is 1.0 inch w.g. (operation to 1.5 inches is permitted), so a person should confirm it.
17The planned final resistance is 1.5 inches w.g.; the recommended final resistance is 1.0 inch w.g. (operation to 1.5 inches is permitted), so a person should confirm it.
18The planned final resistance is 1.6 inches w.g.; the data notes say the filter may be operated to 1.5 inches w.g.
19The operating temperature is 160 F; the maximum operating temperature listed is 158 F.
20The quote states the filter withstands 6.0 inches w.g.; the product sheet lists 5.0 inches w.g. in normal HVAC application.
21The filter is 32 inches deep and the order includes no pocket loops; the product sheet recommends pocket loops for 32 and 36 inch deep filters.
22The filter is 36 inches deep and the order includes no pocket loops; the product sheet recommends pocket loops for 32 and 36 inch deep filters.
23The 20 by 24 header size is available only on consultation with the factory for pricing and availability; its resistance is the same as the listed 24 by 20.
24A filter measuring 20 x 24 x 22 inches with 10 pockets is not a model on the Hi-Flo selection chart.
25The planned final resistance is 1.8 inches w.g.; the data notes say the filter may be operated to 1.5 inches w.g.
26The operating temperature is 170 F; the maximum operating temperature listed is 158 F.
27The quote states the filter withstands 7.0 inches w.g.; the product sheet lists 5.0 inches w.g. in normal HVAC application.
28The design airflow is 3,100 CFM, taken at 100 percent for this system type; the highest listed airflow for this model is 2,800 CFM.
29The quote states 90 square feet of media; the selection chart lists 97 square feet for this model.
30The operating temperature is 200 F; the maximum operating temperature listed is 158 F.
31The quote states an initial resistance of 0.3 inches w.g.; the selection chart lists 0.46 inches w.g. at the rated airflow for this model and efficiency, so a person should confirm it.

  

  

    
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. Model, final resistance and temperature: A Hi-Flo order with an 11-pocket model, 1.8 inch final resistance and 170 F air

    
The agent's request. Please quote a Hi-Flo MERV 13 filter, 24 x 24 x 22 inches with 11 pockets, for a design airflow of 1750 CFM on a constant volume system, planned to a final resistance of 1.8 inches w.g. at 170 F. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Not a listed Hi-Flo model. A filter measuring 24 x 24 x 22 inches with 11 pockets is not a model on the Hi-Flo selection chart.
    • 
      
  • Final resistance above the 1.5 inch limit. The planned final resistance is 1.8 inches w.g.; the data notes say the filter may be operated to 1.5 inches w.g.
    • 
      
  • Operating temperature above the listed maximum. The operating temperature is 170 F; the maximum operating temperature listed is 158 F.
    • 
    

    
2. Model, final resistance and temperature: The agent resubmits after fixing the errors

    
The agent's request. Please quote a Hi-Flo MERV 13 filter, 24 x 24 x 22 inches with 10 pockets, for a design airflow of 1750 CFM on a constant volume system, planned to a final resistance of 1.0 inches w.g. at 150 F. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Listed values for this model. Listed for this model: rated airflow 1,750 CFM (range 1,500 to 2,000 CFM), 73 square feet of media, initial resistance 0.36 inches w.g. at rated airflow.
    • 
    

    
3. Airflow and media area: A deep VAV MERV 14 order at 3,600 CFM with 100 square feet of media

    
The agent's request. Please quote a Hi-Flo MERV 14 filter, 24 x 24 x 36 inches with 8 pockets, for a design airflow of 3600 CFM on a VAV system, planned to a final resistance of 1.0 inches w.g. at 75 F. The quote states 100 square feet of media. Pocket loops are included. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Design airflow above the listed range. The design airflow is 3,600 CFM, taken at 80 percent for this system type; the highest listed airflow for this model is 2,800 CFM.
    • 
      
  • Stated media area differs from the listed area. The quote states 100 square feet of media; the selection chart lists 97 square feet for this model.
    • 
    

    
4. Airflow and media area: The agent resubmits after fixing the errors

    
The agent's request. Please quote a Hi-Flo MERV 14 filter, 24 x 24 x 36 inches with 8 pockets, for a design airflow of 3400 CFM on a VAV system, planned to a final resistance of 1.0 inches w.g. at 75 F. The quote states 97 square feet of media. Pocket loops are included. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Listed values for this model. Listed for this model: rated airflow 2,400 CFM (range 2,000 to 2,800 CFM), 97 square feet of media, initial resistance 0.69 inches w.g. at rated airflow.
    • 
    

    
5. Stated resistance: A MERV 11 order that states an initial resistance of 0.2 inches w.g.

    
The agent's request. Please quote a Hi-Flo MERV 11 filter, 24 x 24 x 30 inches with 10 pockets, for a design airflow of 2400 CFM on a constant volume system, planned to a final resistance of 1.0 inches w.g. at 75 F. The quote states an initial resistance of 0.2 inches w.g.. It is ready to send.

    
Verdict: Needs approval.

    
    
      
  • Stated initial resistance below the listed value. The quote states an initial resistance of 0.2 inches w.g.; the selection chart lists 0.29 inches w.g. at the rated airflow for this model and efficiency, so a person should confirm it.
    • 
      
  • Listed values for this model. Listed for this model: rated airflow 2,400 CFM (range 2,000 to 2,800 CFM), 101 square feet of media, initial resistance 0.29 inches w.g. at rated airflow.
    • 
    

    
6. Stated resistance: The agent resubmits after fixing the errors

    
The agent's request. Please quote a Hi-Flo MERV 11 filter, 24 x 24 x 30 inches with 10 pockets, for a design airflow of 2400 CFM on a constant volume system, planned to a final resistance of 1.0 inches w.g. at 75 F. The quote states an initial resistance of 0.29 inches w.g.. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Listed values for this model. Listed for this model: rated airflow 2,400 CFM (range 2,000 to 2,800 CFM), 101 square feet of media, initial resistance 0.29 inches w.g. at rated airflow.
    • 
    

    
7. Pocket loops: A 36 inch deep Hi-Flo order without pocket loops

    
The agent's request. Please quote a Hi-Flo MERV 13 filter, 24 x 24 x 36 inches with 8 pockets, for a design airflow of 2400 CFM on a constant volume system, planned to a final resistance of 1.0 inches w.g. at 75 F. It is ready to send.

    
Verdict: Needs approval.

    
    
      
  • Pocket loops recommended for 32 and 36 inch filters. The filter is 36 inches deep and the order includes no pocket loops; the product sheet recommends pocket loops for 32 and 36 inch deep filters.
    • 
      
  • Listed values for this model. Listed for this model: rated airflow 2,400 CFM (range 2,000 to 2,800 CFM), 97 square feet of media, initial resistance 0.46 inches w.g. at rated airflow.
    • 
    

    
8. Pocket loops: The agent resubmits after fixing the errors

    
The agent's request. Please quote a Hi-Flo MERV 13 filter, 24 x 24 x 36 inches with 8 pockets, for a design airflow of 2400 CFM on a constant volume system, planned to a final resistance of 1.0 inches w.g. at 75 F. Pocket loops are included. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Listed values for this model. Listed for this model: rated airflow 2,400 CFM (range 2,000 to 2,800 CFM), 97 square feet of media, initial resistance 0.46 inches w.g. at rated airflow.
    • 
    

  

  

    
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 Hi-Flo extended surface pocket filter: the efficiency, the size, the number of pockets, and more. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the published product sheet and returns a verdict, every reason and the listed values it compared against.
    2. 
      
  2. I will start with the first request by the agent.
    3. 
      
  3. The agent says. Please quote a Hi-Flo MERV 13 filter, 24 x 24 x 22 inches with 11 pockets, for a design airflow of 1750 CFM on a constant volume system. It is ready to send. 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. A filter measuring 24 x 24 x 22 inches with 11 pockets is not a model on the Hi-Flo selection chart. The planned final resistance is 1.8 inches w.g.; the data notes say the filter may be operated to 1.5 inches w.g. The operating temperature is 170 F; the maximum operating temperature listed is 158 F.
    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 airflow limits that depend on the system type, a rules-based CPQ can do this only with workarounds.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the pockets to 10, the planned final resistance to 1 and the operating temperature to 150 F.
    10. 
      
  10. It calls the validator again.
    11. 
      
  11. This time the validator clears it.
    12. 
      
  12. All three earlier findings are cleared.
    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.