Predictika · Water treatment resin orders

An AI sales agent's ion exchange resin order, checked before it goes out

An AI sales agent prepares an order for Lewatit ion exchange resin: the grade, the application, the feed conditions, the vessels and bed, and the packs. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of grade suitability, feed limits, bed design and volume tiers, and returns a verdict, every reason, a quote and the figures it computed.

Scope: orders for Lewatit ion exchange resins (MonoPlus S, SP, M, MP, TP grades, CNP 80 and UltraPure MD grades) for softening, demineralization, brine purification, heavy metal removal and ultrapure polishing. See everything this validator covers

This validator models resin orders for a set of Lewatit grades, not just the scenarios shown below. The grade names are LANXESS product names; the rules, limits and prices are modeled from comparable ion exchange resins and are not LANXESS published rules or prices.

  • Grade and application — each of 13 grades is suited to certain applications; a softening grade is not a brine purification grade.
  • Feed limits — the feed temperature and pH must be within the working range of the grade.
  • Bed design — a minimum bed depth for the application, a maximum flow in bed volumes an hour, and enough resin ordered to fill every vessel.
  • Packs and price — 25 litre bags or 1,000 litre big bags (ultrapure grades in bags only), a volume discount by order size, and a total compared with the total the agent states. Very large orders, and orders more than 15 percent above the resin needed, go to a person.

Any Lewatit order that touches these rules is in scope, not only the scenarios pictured here. Not covered: other LANXESS brands, other Lewatit grades, regeneration chemistry and capacity design, freight, taxes, contracts. The quote uses illustrative prices, because LANXESS does not publish a price list; the totals are not LANXESS prices.

The AI sales agentBuilds the resin order from the conversation and says it is ready.
Predictika validatorChecks every rule at once and returns the verdict, each finding, the quote 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 grade that is not made for the applicationOnly with care
A product rule per grade and application. It works, but every new grade is another set of rules to keep in step.		Suitability is looked up for the grade and the application on every call.
Feed temperature and pH limits that differ by gradeOnly with care
A limit table per grade. It works, but the limit has to be compared with a value entered on a different step.		The limits are looked up from the grade and compared with the feed in the same request.
Bed depth and flow limits that depend on the applicationOnly with workarounds
The limits depend on the application and on the vessels and bed entered elsewhere, so a rule needs values from several lines at once.		Vessels, bed, depth and flow are in one request and checked together.
Volume discounts that change at order-size stepsOnly with care
Tiered price rules work, but the step has to be computed from packs and pack size before the discount applies.		The ordered litres, the tier and the total are computed together and compared with the stated total.
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 order; approval findings need a person to sign off; everything else is computed.

    
1This Lewatit grade is not used for the chosen application.
2The feed is at 41 C; this grade is limited to 40 C.
3The feed pH is 10; this grade works between pH 0 and pH 9.
4The bed depth is 699 mm; this application needs at least 700 mm.
5The flow of 81 cubic metres an hour is more than 40 bed volumes an hour for the resin in the vessels (2,000 litres).
6Ultrapure polishing grades are supplied in 25 litre bags only.
7The order is 1,975 litres; the vessels need 2,000 litres of resin.
8The order is 2,325 litres, more than 15 percent above the 2,000 litres the vessels need; a person should confirm the extra resin.
9An order of 20,000 litres (20,000 or more) is priced and scheduled by a person.
10The quote states a total of $1.00; the illustrative price list gives $119,000.00.
11The order needs at least 1 vessel, a bed of at least 1 litre and at least 1 pack.

  

  

    
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 make-up water: anion resin: Anion resin for a 55 C feed in a shallow bed

    
The agent's request. Please quote Lewatit MonoPlus M 600 for boiler make-up demineralization: 4 vessels with a 2,500 litre bed each and a bed depth of 750 mm, a feed at 55 C and pH 8, and a flow of 250 cubic metres an hour. Order 10 x 1,000 litre big bags. The quote states a total of $99,000.00. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Feed too hot for the resin. The feed is at 55 C; this grade is limited to 40 C.
    • 
      
  • Bed too shallow. The bed depth is 750 mm; this application needs at least 800 mm.
    • 
    

    
2. Boiler make-up water: anion resin: The agent resubmits after fixing the errors

    
The agent's request. Please quote Lewatit MonoPlus M 500 for boiler make-up demineralization: 4 vessels with a 2,500 litre bed each and a bed depth of 900 mm, a feed at 55 C and pH 8, and a flow of 250 cubic metres an hour. Order 10 x 1,000 litre big bags. The quote states a total of $108,000.00. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Order summary. 10,000 litres of resin ordered, 10,000 litres needed, list value $120,000.00, volume discount $12,000.00, total $108,000.00 at illustrative prices.
    • 
    

    
3. Brine purification: A softening grade for brine purification with a pH 13 feed and a shallow bed

    
The agent's request. Please quote Lewatit MonoPlus TP 207 for brine purification: 2 vessels with a 1,500 litre bed each and a bed depth of 900 mm, a feed at 60 C and pH 13, and a flow of 40 cubic metres an hour. Order 120 x 25 litre bags. The quote states a total of $68,400.00. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Grade not suited to the application. This Lewatit grade is not used for the chosen application.
    • 
      
  • Feed pH outside the working range. The feed pH is 13; this grade works between pH 2 and pH 12.
    • 
      
  • Bed too shallow. The bed depth is 900 mm; this application needs at least 1,000 mm.
    • 
    

    
4. Brine purification: The agent resubmits after fixing the errors

    
The agent's request. Please quote Lewatit MonoPlus TP 208 for brine purification: 2 vessels with a 1,500 litre bed each and a bed depth of 900 mm, a feed at 60 C and pH 13, and a flow of 40 cubic metres an hour. Order 120 x 25 litre bags. The quote states a total of $91,200.00. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Feed pH outside the working range. The feed pH is 13; this grade works between pH 2 and pH 12.
    • 
      
  • Bed too shallow. The bed depth is 900 mm; this application needs at least 1,000 mm.
    • 
    

    
5. Brine purification: The agent resubmits after fixing the errors

    
The agent's request. Please quote Lewatit MonoPlus TP 260 for brine purification: 2 vessels with a 1,500 litre bed each and a bed depth of 1,000 mm, a feed at 60 C and pH 13, and a flow of 40 cubic metres an hour. Order 120 x 25 litre bags. The quote states a total of $108,300.00. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Order summary. 3,000 litres of resin ordered, 3,000 litres needed, list value $114,000.00, volume discount $5,700.00, total $108,300.00 at illustrative prices.
    • 
    

    
6. Ultrapure water polishing: Ultrapure polishing resin in big bags for a 20,000 litre order

    
The agent's request. Please quote Lewatit UltraPure 1294 MD for ultrapure water polishing: 4 vessels with a 5,000 litre bed each and a bed depth of 1,200 mm, a feed at 25 C and pH 7, and a flow of 800 cubic metres an hour. Order 20 x 1,000 litre big bags. The quote states a total of $340,000.00. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Pack size not offered for this grade. Ultrapure polishing grades are supplied in 25 litre bags only.
    • 
      
  • Large volume order to confirm. An order of 20,000 litres (20,000 or more) is priced and scheduled by a person.
    • 
    

    
7. Ultrapure water polishing: The agent resubmits after fixing the errors

    
The agent's request. Please quote Lewatit UltraPure 1294 MD for ultrapure water polishing: 4 vessels with a 5,000 litre bed each and a bed depth of 1,200 mm, a feed at 25 C and pH 7, and a flow of 800 cubic metres an hour. Order 800 x 25 litre bags. The quote states a total of $340,000.00. It is ready to send.

    
Verdict: Needs approval.

    
    
      
  • Large volume order to confirm. An order of 20,000 litres (20,000 or more) is priced and scheduled by a person.
    • 
      
  • Order summary. 20,000 litres of resin ordered, 20,000 litres needed, list value $400,000.00, volume discount $60,000.00, total $340,000.00 at illustrative prices.
    • 
    

  

  

    
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 Lewatit ion exchange resin: the grade, the application, the feed conditions, and more. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of grade suitability, feed limits, bed design and volume tiers, and returns a verdict, every reason, a quote and the figures it computed.
    2. 
      
  2. I will start with the first request by the agent.
    3. 
      
  3. The agent says. Please quote Lewatit MonoPlus M 600 for boiler make-up demineralization: 4 vessels. Order 10 x 1,000 litre big bags. The rest of the request is on screen. The quote states a total of $99,000.00.
    4. 
      
  4. Before it goes out, the agent calls the validator.
    5. 
      
  5. Blocked, with two findings, all in one answer. The feed is at 55 C; this grade is limited to 40 C. The bed depth is 750 mm; this application needs at least 800 mm.
    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 order are checked against each other.
    7. 
      
  7. What this shows: for feed temperature and pH limits that differ by grade, a rules-based CPQ is error-prone for this.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the resin grade to Lewatit MonoPlus M 500, the bed depth to 900 mm and the stated total to $108,000.00.
    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, on screen.
    13. 
      
  13. Both earlier findings are cleared, and the validator's total, $108,000.00, matches the order.
    14. 
      
  14. Two problems caught in one call, fixed in one pass, and a clean $108,000.00 order. That is the value of Predictika's validator: catching mistakes before they reach the customer. Now you can try the other scenarios yourself.
    15.