Predictika · Paper machine clothing orders

An AI sales agent's paper machine clothing order, checked before it goes out

An AI sales agent prepares an order for paper machine clothing: forming fabrics, press felts, seamed press fabrics, dryer fabrics or process belts for a tissue, packaging or publication machine, with the machine section, the dimensions, the machine speed, the lead time, the service level, the number of items and a discount. Before the order goes out, the agent sends it to Predictika, which checks it against a declarative model of the product line and returns a verdict, every reason and the total it computed.

Scope: Forming fabrics, press felts, seamed press fabrics, dryer fabrics and process belts: product against machine section and grade, dimensions, speed, lead time, service level, discount tiers and the stated total. See everything this validator covers

This validator models paper machine clothing orders, not just the scenarios shown below.

  • Product and machine section — each product belongs in certain sections of the machine, forming fabrics are multilayer with one to three layers, and shoe press belts are quoted for publication and packaging grades, not tissue.
  • Dimensions, speed and lead time — width and length ranges, a machine speed limit with a lower one for single-layer forming fabrics, and a minimum lead time for each product.
  • Service levels — Regular, Consultative, Diagnostic and Project Specific service, where Diagnostic service is a long-term analysis that needs an order of at least two items.
  • Approvals and totals — a discount above the tier for the order size, a process belt, a fast machine and a very wide machine need a person, a discount above 20 percent is blocked, and the total the order states is checked against the computed one.

Any paper machine clothing order that touches these rules is in scope — not only the scenarios pictured here. Not covered: Engineered Fabrics outside paper, composites, other manufacturers' products, installation and negotiated contract pricing. Product categories, service levels and grade segments follow public Albany International pages and filings. All prices in US dollars, the discount tiers, dimension, speed and lead-time limits and the product and section pairing are modeled from comparable products and are illustrative, not published by Albany International.

The AI sales agentBuilds the clothing order from the conversation with the paper mill 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
A product that must suit the machine section and gradeOnly with care
Each product is valid in certain sections, and shoe press belts are excluded for one grade, so the pairing is a rule on three fields.
The product, the machine section and the grade are fields of one order and are checked together on every call.
A speed limit that depends on the layer countOnly with care
Single-layer forming fabrics have a lower speed limit than the general limit, so the speed is compared twice against different values.
The layer count and the speed are read together and the finding names which limit was exceeded.
A service level that needs a minimum order sizeOnly with workarounds
Diagnostic service is only offered on an order of two or more items, so one choice constrains another.
Both choices are fields of the order and are compared in the same pass.
A discount whose limit depends on the order sizeOnly with workarounds
The discount that needs approval depends on the number of items, so the limit has to be looked up before it is compared.
The tier is computed inside the model and compared with the discount on every call.
A stated total that must match the computed oneOnly with workarounds
The price depends on the product, layers, dimensions, number of items, service fee and discount, so the total has to be recomputed before it is compared.
The total is computed inside the model and compared with the stated one on every call.
Every problem returned in one answerOnly with workarounds
Rules tend to fire as the order is saved or a step is completed, so a person often fixes one error and meets the next.
All findings come back together, so the agent can fix them in a single pass.

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.

1Forming fabrics are multilayer with 1 to 3 layers and other products have no layer count; the quote has 4 layers.
2A single layer forming fabric is limited to 600 m/min; the quote has 800 m/min.
3The selected product is not used in the selected machine section.
4A process belt is selected for the specific project; an application engineer should confirm the design.
5Shoe press belts are quoted for publication and packaging grades; tissue machines use press felts and seamed press fabrics.
6At 1,800 m/min a fabric specialist should confirm the design.
7A machine speed of 2,500 m/min is above the 2,200 m/min the model supports.
8A width of 1,200 cm is outside the 150 to 1,100 cm range.
9A width of 1,000 cm needs a fabric specialist to confirm handling and installation.
10A length of 70 m is outside the 5 to 60 m range.
11The requested lead time of 4 weeks is shorter than the minimum for this product.
12Diagnostic service is a long-term analysis and needs an order of at least 2 items.
13The order has 0 fabrics or belts; at least 1 is needed.
14A discount of 8 percent is above the 5 percent allowed for this order size; a sales manager should approve it.
15A discount of 25 percent is above the 20 percent limit.
16The quote states a total of USD 10; the modeled prices give USD 166,400.

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. Forming fabric for the press section: Forming fabric ordered for the wrong section, single layer, too wide, with a short lead time

The agent's request. Please quote 4 forming fabrics for the press section of a packaging machine, with 1 layer. The machine is 1,250 cm wide with a 40 m length and runs at 1,200 m/min. The customer wants delivery in 3 weeks and Consultative service. The sales rep offers a discount of 7 percent and the order states a total of USD 334,800. It is ready to send.

Verdict: Blocked.

  • Product not used in this machine section. The selected product is not used in the selected machine section.
  • Single layer forming fabric too fast. A single layer forming fabric is limited to 600 m/min; the quote has 1,200 m/min.
  • Machine width outside range. A width of 1,250 cm is outside the 150 to 1,100 cm range.
  • Lead time too short. The requested lead time of 3 weeks is shorter than the minimum for this product.
  • Stated total does not match. The quote states a total of USD 334,800; the modeled prices give USD 336,300.

2. Forming fabric for the press section: The agent resubmits after fixing the errors

The agent's request. Please quote 4 forming fabrics for the forming section of a packaging machine, with 2 layers. The machine is 1,000 cm wide with a 40 m length and runs at 1,200 m/min. The customer wants delivery in 5 weeks and Consultative service. The sales rep offers a discount of 7 percent and the order states a total of USD 313,980. It is ready to send.

Verdict: Needs approval.

  • Very wide machine to confirm. A width of 1,000 cm needs a fabric specialist to confirm handling and installation.
  • Order summary. 4 items at USD 84,000 each, 7 percent discount, modeled, illustrative total of USD 313,980.

3. Forming fabric for the press section: The agent resubmits after fixing the errors

The agent's request. Please quote 4 forming fabrics for the forming section of a packaging machine, with 2 layers. The machine is 880 cm wide with a 40 m length and runs at 1,200 m/min. The customer wants delivery in 5 weeks and Consultative service. The sales rep offers a discount of 7 percent and the order states a total of USD 276,482.40. It is ready to send.

Verdict: Allowed.

  • Order summary. 4 items at USD 73,920 each, 7 percent discount, modeled, illustrative total of USD 276,482.40.

4. Shoe press belt for a tissue machine: Shoe press belt for a tissue machine with Diagnostic service on one item

The agent's request. Please quote 1 shoe press belt for the press section of a tissue machine. The machine is 1,000 cm wide with a 40 m length and runs at 1,800 m/min. The customer wants delivery in 8 weeks and Diagnostic service. The sales rep offers a discount of 10 percent and the order states a total of USD 328,000. It is ready to send.

Verdict: Blocked.

  • Shoe press belt not for tissue. Shoe press belts are quoted for publication and packaging grades; tissue machines use press felts and seamed press fabrics.
  • Diagnostic service needs a multi-item order. Diagnostic service is a long-term analysis and needs an order of at least 2 items.
  • Lead time too short. The requested lead time of 8 weeks is shorter than the minimum for this product.
  • Discount above the volume tier. A discount of 10 percent is above the 5 percent allowed for this order size; a sales manager should approve it.
  • Process belt needs application review. A process belt is selected for the specific project; an application engineer should confirm the design.
  • Fast machine to confirm. At 1,800 m/min a fabric specialist should confirm the design.
  • Very wide machine to confirm. A width of 1,000 cm needs a fabric specialist to confirm handling and installation.

5. Shoe press belt for a tissue machine: The agent resubmits after fixing the errors

The agent's request. Please quote 1 seamed press fabric for the press section of a tissue machine. The machine is 1,000 cm wide with a 40 m length and runs at 1,800 m/min. The customer wants delivery in 8 weeks and Diagnostic service. The sales rep offers a discount of 10 percent and the order states a total of USD 115,600. It is ready to send.

Verdict: Blocked.

  • Diagnostic service needs a multi-item order. Diagnostic service is a long-term analysis and needs an order of at least 2 items.
  • Discount above the volume tier. A discount of 10 percent is above the 5 percent allowed for this order size; a sales manager should approve it.
  • Fast machine to confirm. At 1,800 m/min a fabric specialist should confirm the design.
  • Very wide machine to confirm. A width of 1,000 cm needs a fabric specialist to confirm handling and installation.

6. Shoe press belt for a tissue machine: The agent resubmits after fixing the errors

The agent's request. Please quote 3 seamed press fabrics for the press section of a tissue machine. The machine is 850 cm wide with a 40 m length and runs at 1,200 m/min. The customer wants delivery in 8 weeks and Diagnostic service. The sales rep offers a discount of 5 percent and the order states a total of USD 304,390. It is ready to send.

Verdict: Allowed.

  • Order summary. 3 items at USD 105,400 each, 5 percent discount, modeled, illustrative total of USD 304,390.

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 paper machine clothing: forming fabrics, press felts, seamed press fabrics, and more. Before the order goes out, the agent sends it to Predictika, which checks it against a declarative model of the product line and returns a verdict, every reason and the total it computed.
  2. I will start with the first request by the agent.
  3. The agent says. Please quote 4 forming fabrics for the press section of a packaging machine, with 1 layer. The machine is 1,250 cm wide with a 40 m length and runs at 1,200 m/min. The rest of the request is on screen. The sales rep offers a discount of 7 percent and the order states a total of USD 334,800.
  4. Before it goes out, the agent calls the validator.
  5. Blocked, with five findings, all in one answer. The selected product is not used in the selected machine section. A single layer forming fabric is limited to 600 m/min; the quote has 1,200 m/min. A width of 1,250 cm is outside the 150 to 1,100 cm range. The other two are on screen.
  6. The validator returned five 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 and against the stated total.
  7. What this shows: for a product that must suit the machine section and grade, a rules-based CPQ is error-prone for this.
  8. Now the agent resubmits.
  9. The corrected request. It changes the machine section to forming, the layers to 2, the width to 1,000 cm, the lead time to 5 weeks and the total stated on the order to USD 313,980.
  10. It calls the validator again.
  11. Not blocked, but held for a person.
  12. The five blocking findings are cleared; one approval remains for a person: very wide machine to confirm.
  13. The agent resubmits once more.
  14. The revised request. It changes the width to 880 cm and the total stated on the order to USD 276,482.40.
  15. It calls the validator once more.
  16. This time the validator clears it.
  17. The earlier finding is cleared.
  18. Five problems caught in one call, fixed in two 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.