Predictika · Emissions source testing quotes

An AI sales agent's emissions source test quote, checked before it goes out

An AI sales agent prepares a quote for an air emissions source testing program: compliance stack test, CEMS relative accuracy test or engineering and diagnostics test, the EPA methods, number of runs and run length, unit load, notice to the agency, crew days, report turnaround, travel and a total. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of source testing practice and returns a verdict, every reason and the figures it computed.

Scope: Montrose-style air emissions source testing: compliance stack tests, CEMS relative accuracy tests and engineering and diagnostics tests, with EPA Methods 3A, 5, 7E, 10, 25A, 29 and 320*. See everything this validator covers

This validator models quotes for three source testing services of the kind Montrose Air Quality Services describes publicly, not just the scenarios shown below. Items marked with an asterisk are illustrative.

  • Services — compliance stack testing, CEMS testing (relative accuracy test audits) and engineering and diagnostics testing; each has its own minimum number of runs, run length, unit load and notice to the agency.
  • Methods — EPA Methods 3A (oxygen and carbon dioxide), 7E (nitrogen oxides), 10 (carbon monoxide), 25A (volatile organic compounds), 5 (filterable particulate) and 29 (metals), plus Method 320 (FTIR multi-gas)*; CEMS tests accept only the gas analyzer methods.
  • Source and program — boiler, turbine, engine, thermal oxidizer or process source; units, sampling locations, crews (1 to 3); a destruction efficiency test* for a thermal oxidizer needs an inlet and an outlet location and Method 25A.
  • Schedule and money — crew days worked out from runs, run length, units and crews; report turnaround (a rush schedule needs approval, under 10 days is blocked); mileage and per diem for distant sites; a discount that needs approval above 15 percent of the subtotal and is blocked above 30 percent; the agent's stated days and total are compared with the computed ones.

Any quote for these three services that touches these rules is in scope — not only the scenarios pictured here. Not covered: ambient air monitoring, leak detection surveys, laboratory analysis, assessment and permitting, remediation, other Montrose services, and state-specific permit conditions. Rates and thresholds are typical for source testing and are not published by Montrose; the footer says which items are real and which are illustrative.

The AI sales agentBuilds the source test quote 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
Run count and run length that depend on the service and the methods (9 runs for a CEMS test, 120-minute runs for metals)Only with care
Each service and method pairing is its own comparison rule; a rep has to remember that metals need longer runs than gases and that a CEMS test needs more runs, and every new method adds a rule.		Service, methods, runs and run length are checked together on every call.
Crew days worked out from runs, run length, units and crews, and compared with the days the agent statesOnly with workarounds
The day count is a stepped calculation (sampling hours per crew per day, rounded up, plus a setup day) that a rules-based CPQ handles with helper fields, and a stated figure is a free-entry field that someone has to compare by hand.		The days are computed from the program, and the stated days and total are compared with them on every call.
Notice to the agency, a submitted protocol and the unit load, each different by serviceOnly with care
Three requirements with different thresholds for each service; they can be written as rules, but each depends on the service chosen and is easy to leave out.		Each requirement is checked for the service chosen and every miss is reported together.
Every problem returned in one answerOnly with workarounds
Rules tend to fire as the quote is saved or a step is completed, so a rep often fixes one error and then 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 program has 0 unit(s), 1 sampling location(s) and 1 crew(s); it needs at least 1 unit, 1 location and 1 to 3 crews.
2The program has 1 unit(s), 0 sampling location(s) and 1 crew(s); it needs at least 1 unit, 1 location and 1 to 3 crews.
3The program has 1 unit(s), 1 sampling location(s) and 4 crew(s); it needs at least 1 unit, 1 location and 1 to 3 crews.
4No EPA measurement method is selected for the program.
5A CEMS relative accuracy test is run against the gas analyzers (Methods 3A, 7E, 10 and 25A); particulate, metals and FTIR methods are not part of it.
6The program has 2 run(s) per unit; this service needs at least 3.
7The program has 8 run(s) per unit; this service needs at least 9.
8The program has 0 run(s) per unit; this service needs at least 1.
9Runs of 59 minutes are shorter than the 60 minutes this service and method set needs.
10Runs of 119 minutes are shorter than the 120 minutes this service and method set needs.
11Runs of 20 minutes are shorter than the 21 minutes this service and method set needs.
12Runs of 19 minutes are shorter than the 20 minutes this service and method set needs.
13The unit would be tested at 89 percent load; this service needs at least 90 percent.
14The unit would be tested at 49 percent load; this service needs at least 50 percent.
1529 days of notice before the test date is less than the 30 days this service needs.
1620 days of notice before the test date is less than the 21 days this service needs.
17The test protocol has not been submitted to the agency, which this service needs before the test.
18A destruction efficiency test is for a thermal oxidizer or similar control device and uses Method 25A for VOC.
19A destruction efficiency test needs an inlet and an outlet location; the program has 1.
20A final report in 9 days is shorter than the 10 days of the fastest rush schedule.
21A final report in 10 days is a rush schedule with a surcharge; a person should confirm lab and report staff.
22A final report in 19 days is a rush schedule with a surcharge; a person should confirm lab and report staff.
23A discount of $2,500 is more than 15 percent of the $14,750 subtotal; a person should confirm.
24A discount of $4,425 is more than 15 percent of the $14,750 subtotal; a person should confirm.
25A discount of $4,500 is more than 30 percent of the $14,750 subtotal.
26The quote states 1 crew day(s) of work; the program needs 2 crew day(s) (one setup day plus the sampling days).
27The quote states a total of $20,000; the typical rates give $14,990.
28The program needs 11 crew days on site; a person should confirm crew scheduling.
29The site is 301 miles from the office; a person should confirm travel and per diem.
30A program on 4 units is a large project; a person should confirm project management.
3110 days of notice before the test date is less than the 30 days this service needs.
32A final report in 5 days is shorter than the 10 days of the fastest rush schedule.

  

  

    
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. Boiler compliance test: A two-boiler compliance test with particulate and metals, short runs and a low load

    
The agent's request. Please quote a compliance stack test on 2 boilers with Method 7E (NOx), Method 10 (CO), Method 3A (O2 and CO2), Method 5 (filterable particulate) and Method 29 (metals), 3 runs of 60 minutes per unit, 1 sampling location and 1 crew. The units will run at 80 percent load, the agency will get 20 days of notice, and the test protocol has not been submitted. The site is 120 miles away and the customer wants the final report in 30 days. The quote states 2 crew days and a total of $31,460. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Test runs too short. Runs of 60 minutes are shorter than the 120 minutes this service and method set needs.
    • 
      
  • Unit load too low for the test. The unit would be tested at 80 percent load; this service needs at least 90 percent.
    • 
      
  • Notice to the agency too short. 20 days of notice before the test date is less than the 30 days this service needs.
    • 
      
  • Test protocol not submitted. The test protocol has not been submitted to the agency, which this service needs before the test.
    • 
    

    
2. Boiler compliance test: The agent resubmits after fixing the errors

    
The agent's request. Please quote a compliance stack test on 2 boilers with Method 7E (NOx), Method 10 (CO), Method 3A (O2 and CO2), Method 5 (filterable particulate) and Method 29 (metals), 3 runs of 120 minutes per unit, 1 sampling location and 1 crew. The units will run at 95 percent load, the agency will get 45 days of notice, and the test protocol has been submitted. The site is 120 miles away and the customer wants the final report in 30 days. The quote states 3 crew days and a total of $36,200. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Program summary. 3 crew day(s) on site, estimated total of $36,200 at typical rates.
    • 
    

    
3. CEMS test on a turbine: A CEMS relative accuracy test with three runs and a particulate method

    
The agent's request. Please quote a CEMS relative accuracy test on 1 turbine with Method 7E (NOx), Method 3A (O2 and CO2) and Method 5 (filterable particulate), 3 runs of 21 minutes per unit, 1 sampling location and 1 crew. The unit will run at 70 percent load, the agency will get 30 days of notice, and the test protocol has been submitted. The site is 90 miles away and the customer wants the final report in 15 days. The quote states 2 crew days and a total of $18,360. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Method not used for CEMS relative accuracy tests. A CEMS relative accuracy test is run against the gas analyzers (Methods 3A, 7E, 10 and 25A); particulate, metals and FTIR methods are not part of it.
    • 
      
  • Too few test runs. The program has 3 run(s) per unit; this service needs at least 9.
    • 
      
  • Rush report schedule. A final report in 15 days is a rush schedule with a surcharge; a person should confirm lab and report staff.
    • 
    

    
4. CEMS test on a turbine: The agent resubmits after fixing the errors

    
The agent's request. Please quote a CEMS relative accuracy test on 1 turbine with Method 7E (NOx) and Method 3A (O2 and CO2), 9 runs of 21 minutes per unit, 1 sampling location and 1 crew. The unit will run at 70 percent load, the agency will get 30 days of notice, and the test protocol has been submitted. The site is 90 miles away and the customer wants the final report in 15 days. The quote states 2 crew days and a total of $15,810. It is ready to send.

    
Verdict: Needs approval.

    
    
      
  • Rush report schedule. A final report in 15 days is a rush schedule with a surcharge; a person should confirm lab and report staff.
    • 
      
  • Program summary. 2 crew day(s) on site, estimated total of $15,810 at typical rates.
    • 
    

    
5. CEMS test on a turbine: The agent resubmits after fixing the errors

    
The agent's request. Please quote a CEMS relative accuracy test on 1 turbine with Method 7E (NOx) and Method 3A (O2 and CO2), 9 runs of 21 minutes per unit, 1 sampling location and 1 crew. The unit will run at 70 percent load, the agency will get 30 days of notice, and the test protocol has been submitted. The site is 90 miles away and the customer wants the final report in 30 days. The quote states 2 crew days and a total of $12,810. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Program summary. 2 crew day(s) on site, estimated total of $12,810 at typical rates.
    • 
    

    
6. Thermal oxidizer: A thermal oxidizer destruction efficiency test with one location and a large discount

    
The agent's request. Please quote a compliance stack test on 1 thermal oxidizer with Method 25A (VOC), as a destruction efficiency test, 3 runs of 60 minutes per unit, 1 sampling location and 1 crew. The unit will run at 95 percent load, the agency will get 40 days of notice, and the test protocol has been submitted. The site is 60 miles away and the customer wants the final report in 30 days, with a discount of $6,500. The quote states 2 crew days and a total of $7,740. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Destruction efficiency needs inlet and outlet. A destruction efficiency test needs an inlet and an outlet location; the program has 1.
    • 
      
  • Discount above 30 percent of the subtotal. A discount of $6,500 is more than 30 percent of the $14,000 subtotal.
    • 
    

    
7. Thermal oxidizer: The agent resubmits after fixing the errors

    
The agent's request. Please quote a compliance stack test on 1 thermal oxidizer with Method 25A (VOC), as a destruction efficiency test, 3 runs of 60 minutes per unit, 2 sampling locations and 1 crew. The unit will run at 95 percent load, the agency will get 40 days of notice, and the test protocol has been submitted. The site is 60 miles away and the customer wants the final report in 30 days, with a discount of $1,800. The quote states 2 crew days and a total of $12,440. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Program summary. 2 crew day(s) on site, estimated total of $12,440 at typical rates.
    • 
    

  

  

    
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 quote for an air emissions source testing program: compliance stack test, CEMS relative accuracy test or engineering and diagnostics test, the EPA methods, and more. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of source testing practice 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 a compliance stack test on 2 boilers with Method 7E (NOx), Method 10 (CO), Method 3A (O2 and CO2). The units will run at 80 percent load, the agency will get 20 days of notice, and the test protocol has not been submitted. The rest of the request is on screen. The quote states 2 crew days and a total of $31,460.
    4. 
      
  4. Before it goes out, the agent calls the validator.
    5. 
      
  5. Blocked, with four findings, all in one answer. Runs of 60 minutes are shorter than the 120 minutes this service and method set needs. The unit would be tested at 80 percent load; this service needs at least 90 percent. 20 days of notice before the test date is less than the 30 days this service needs. The other one is on screen.
    6. 
      
  6. The validator returned four 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 notice to the agency, a submitted protocol and the unit load, each different by service, a rules-based CPQ is error-prone for this.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the minutes per run to 120, the unit load to 95 percent, the notice to the agency to 45 days, the protocol submitted to yes, the stated crew days to 3 and the stated total to $36,200.
    10. 
      
  10. It calls the validator again.
    11. 
      
  11. This time the validator clears it.
    12. 
      
  12. The validator also returns the line-item quote. It starts with Compliance stack test crew-day on site (2 test days plus a setup day and Method 7E (NOx); the other nine lines are on screen.
    13. 
      
  13. All four earlier findings are cleared, and the validator's total, $36,200, matches the quote.
    14. 
      
  14. Four problems caught in one call, fixed in one pass, and a clean $36,200 quote. That is the value of Predictika's validator: catching mistakes before they reach the customer. Now you can try the other scenarios yourself.
    15.