Predictika · Melting unit orders

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

An AI sales agent prepares an order for a scrap and DRI melting unit: the furnace size, the heats per year, the tap size the plant needs, the charge mix, the ECS continuous charging system and Hytemp, the Q-One or conventional power feeder with SVC, Q-Melt process control, the grid short-circuit power and the annual output the order states. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of how these choices fit together, and returns a verdict, every reason and the figures it computed.

Scope: One melting unit order: furnace size, heats, charge mix, charging system, power feeder, Q-Melt, grid strength and the stated annual output. See everything this validator covers

This validator models four furnace sizes and their power and charging options, not just the scenarios shown below.

  • Size and output — four Digimelter sizes with their tap size and power rating, the limit of 15,000 heats a year, and the annual output from heats and tap size.
  • Charging — scrap, scrap with DRI or DRI only, the ECS continuous charging system that DRI needs and the Hytemp hot-DRI transport a DRI-only charge needs.
  • Power — the Q-One power feeder or a conventional feeder that needs SVC grid compensation, Q-Melt process control, and the grid short-circuit power each combination requires.
  • Approvals — a thin grid margin that needs a grid study and a DRI-only process that needs a process review.

Any melting unit order that touches these rules is in scope — not only the scenarios pictured here. Not covered: other Danieli equipment and prices. The Digimelter, Q-One, Q-Melt, ECS and Hytemp names and the 15,000 heats limit come from public Danieli pages; Danieli does not publish its size ladder, power ratings or grid requirements, so those are modeled from comparable furnaces and are not published by Danieli. No prices are modeled.

The AI sales agentBuilds the order from the conversation with the customer and says it is ready to send.
Predictika validatorChecks every rule at once and returns the verdict, each finding, and the computed figures.

6 scenarios in 2 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
Charging systems tied to the charge mixOnly with workarounds
A DRI charge needs continuous charging and a DRI-only charge also needs hot transport, so the same choice is required by one mix and not by another.		Charge mix and charging systems are fields of one order and are checked against each other on every call.
Power feeder and grid strengthOnly with care
The grid strength required depends on the furnace size and on the feeder, and a conventional feeder also needs SVC compensation.		Size, feeder, SVC and grid are compared together on every call.
Options available with only one feederOnly with care
Q-Melt is offered with the Q-One feeder only, a dependency between two separate choices.		The dependency is checked on every call, in either direction.
Blocks and approvals at onceOnly with workarounds
A thin grid margin and a DRI-only process need a person while a missing system stops the order, and both should be seen together.		Blocks and approvals are reported together, so one problem never hides another.
A stated output that must match the computed oneOnly with workarounds
Annual output is heats times the tap size of the chosen furnace, so a separate rule must recompute it and compare it with the stated figure.		The stated output is a field of the order and is compared with the computed one on every call.

  

  

    
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.

    
1The order asks for 15,001 heats per year; the Digimelter is quoted for up to 15,000.
2The customer needs 121 tons per heat; this Digimelter size taps 120.
3The order states 900,000 t of steel per year; heats times tap size give 840,000 t.
4DRI in the charge mix needs the ECS continuous charging system.
5A DRI-only charge needs a process review of chemistry and slag practice.
6A DRI-only charge needs the Hytemp hot-DRI transport system.
7A conventional furnace transformer needs SVC grid compensation; the Q-One feeder does not.
8The Q-Melt Melt-Model works with the Q-One electrical control; it is not offered with a conventional feeder.
9The grid short-circuit power (500 MVA) is less than twice the minimum (500 MVA); a grid study should confirm.
10The grid short-circuit power is 499 MVA; this furnace and feeder need at least 500 MVA.

  

  

    
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. DRI melting unit order: A 90-ton Digimelter for 100-ton heats, 16,000 heats a year, DRI only, on a 300 MVA grid

    
The agent's request. Please quote a 90-ton Digimelter for a plant that needs 100 tons per heat, 16,000 heats per year, a DRI only charge, the Q-One power feeder and Q-Melt, on a grid with 300 MVA of short-circuit power. The order states 1,440,000 tons of steel per year. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Heats per year over 15,000. The order asks for 16,000 heats per year; the Digimelter is quoted for up to 15,000.
    • 
      
  • Tap size too small. The customer needs 100 tons per heat; this Digimelter size taps 90.
    • 
      
  • DRI charge needs continuous charging. DRI in the charge mix needs the ECS continuous charging system.
    • 
      
  • DRI-only charge needs Hytemp. A DRI-only charge needs the Hytemp hot-DRI transport system.
    • 
      
  • Grid too weak for the furnace. The grid short-circuit power is 300 MVA; this furnace and feeder need at least 375 MVA.
    • 
      
  • DRI-only process review. A DRI-only charge needs a process review of chemistry and slag practice.
    • 
    

    
2. DRI melting unit order: The agent resubmits after fixing the errors

    
The agent's request. Please quote a 120-ton Digimelter for a plant that needs 100 tons per heat, 15,000 heats per year, a DRI only charge, with the ECS continuous charging system, the Q-One power feeder and Q-Melt, on a grid with 300 MVA of short-circuit power. The order states 1,800,000 tons of steel per year. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • DRI-only charge needs Hytemp. A DRI-only charge needs the Hytemp hot-DRI transport system.
    • 
      
  • Grid too weak for the furnace. The grid short-circuit power is 300 MVA; this furnace and feeder need at least 500 MVA.
    • 
      
  • DRI-only process review. A DRI-only charge needs a process review of chemistry and slag practice.
    • 
    

    
3. DRI melting unit order: The agent resubmits after fixing the errors

    
The agent's request. Please quote a 120-ton Digimelter for a plant that needs 100 tons per heat, 15,000 heats per year, a DRI only charge, with the ECS continuous charging system and Hytemp, the Q-One power feeder and Q-Melt, on a grid with 700 MVA of short-circuit power. The order states 1,800,000 tons of steel per year. It is ready to send.

    
Verdict: Needs approval.

    
    
      
  • Thin grid margin. The grid short-circuit power (700 MVA) is less than twice the minimum (500 MVA); a grid study should confirm.
    • 
      
  • DRI-only process review. A DRI-only charge needs a process review of chemistry and slag practice.
    • 
      
  • Furnace summary. 120 t tap, 100 MVA, 15,000 heats per year, 1,800,000 t of steel per year.
    • 
    

    
4. DRI melting unit order: The agent resubmits after fixing the errors

    
The agent's request. Please quote a 120-ton Digimelter for a plant that needs 100 tons per heat, 15,000 heats per year, a scrap and DRI charge, with the ECS continuous charging system and Hytemp, the Q-One power feeder and Q-Melt, on a grid with 1,200 MVA of short-circuit power. The order states 1,800,000 tons of steel per year. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Furnace summary. 120 t tap, 100 MVA, 15,000 heats per year, 1,800,000 t of steel per year.
    • 
    

    
5. Conventional feeder order: A 150-ton Digimelter on a conventional feeder with Q-Melt and no SVC

    
The agent's request. Please quote a 150-ton Digimelter for a plant that needs 150 tons per heat, 8,000 heats per year, a scrap only charge, with the ECS continuous charging system, a conventional power feeder and Q-Melt, on a grid with 2,500 MVA of short-circuit power. The order states 1,000,000 tons of steel per year. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Stated annual output does not match. The order states 1,000,000 t of steel per year; heats times tap size give 1,200,000 t.
    • 
      
  • Conventional feeder needs SVC. A conventional furnace transformer needs SVC grid compensation; the Q-One feeder does not.
    • 
      
  • Q-Melt needs the Q-One feeder. The Q-Melt Melt-Model works with the Q-One electrical control; it is not offered with a conventional feeder.
    • 
    

    
6. Conventional feeder order: The agent resubmits after fixing the errors

    
The agent's request. Please quote a 150-ton Digimelter for a plant that needs 150 tons per heat, 8,000 heats per year, a scrap only charge, with the ECS continuous charging system, a conventional power feeder with SVC and no Q-Melt, on a grid with 2,500 MVA of short-circuit power. The order states 1,200,000 tons of steel per year. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Furnace summary. 150 t tap, 125 MVA, 8000 heats per year, 1,200,000 t of steel per year.
    • 
    

  

  

    
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 scrap and DRI melting unit: the furnace size, the heats per year, the tap size the plant needs, and more. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of how these choices fit together, 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 90-ton Digimelter for a plant that needs 100 tons per heat, 16,000 heats per year, a DRI only charge. The order states 1,440,000 tons of steel per year. The rest of the request is on screen.
    4. 
      
  4. Before it goes out, the agent calls the validator.
    5. 
      
  5. Blocked, with five findings and one approval, all in one answer. The order asks for 16,000 heats per year; the Digimelter is quoted for up to 15,000. The customer needs 100 tons per heat; this Digimelter size taps 90. DRI in the charge mix needs the ECS continuous charging system. The other two are on screen.
    6. 
      
  6. The validator returned five blocking findings and one approval 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 charging systems tied to the charge mix, a rules-based CPQ can do this only with workarounds.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the furnace size to 120-ton, the heats per year to 15,000, the ECS continuous charging to yes and the stated annual output to 1,800,000.
    10. 
      
  10. It calls the validator again.
    11. 
      
  11. Still blocked, with two findings and one approval, all in one answer. A DRI-only charge needs the Hytemp hot-DRI transport system. The grid short-circuit power is 300 MVA; this furnace and feeder need at least 500 MVA.
    12. 
      
  12. Three findings cleared. Still blocking: DRI-only charge needs Hytemp and grid too weak for the furnace. One approval also remains for a person: DRI-only process review.
    13. 
      
  13. The agent resubmits once more.
    14. 
      
  14. The revised request. It changes the hytemp to yes and the grid short-circuit power to 700.
    15. 
      
  15. It calls the validator once more.
    16. 
      
  16. Not blocked, but held for a person.
    17. 
      
  17. The two blocking findings are cleared; two approvals remain for a person: thin grid margin and DRI-only process review.
    18. 
      
  18. The agent resubmits again.
    19. 
      
  19. The next revised request. It changes the charge mix to scrap and DRI and the grid short-circuit power to 1,200.
    20. 
      
  20. It calls the validator yet again.
    21. 
      
  21. This time the validator clears it.
    22. 
      
  22. Both earlier findings are cleared.
    23. 
      
  23. Six problems caught in one call, fixed in three passes, and a clean order. That is the value of Predictika's validator: catching mistakes before they reach the customer. Now you can try the other scenarios yourself.
    24.