Predictika · Semiconductor subsystem quotes

An AI sales agent's gas and chemical delivery subsystem quote, checked before it goes out

An AI sales agent prepares a quote for a gas and chemical delivery subsystem: a process tool gas panel, a valve manifold or an all-plastic chemical delivery module, with a material, a service type, a cleanroom class, heaters, a leak test, a quantity, a lead time and a discount. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the product rules and returns a verdict, every reason and the total it computed.

Scope: Gas panel, valve manifold and chemical delivery module quotes: material and product fit, channel limits, service and leak-test rules, cleanroom class, lead time, discount approval and the stated total. See everything this validator covers

This validator models gas and chemical delivery subsystem quotes, not just the scenarios shown below.

  • Product, material and cleanroom — stainless steel for gas panels and valve manifolds, plastics for chemical delivery modules, the channel range of each product, and the Class 100, Class 1000 and ISO 7 build options.
  • Service and tests — corrosive service needs electropolished steel, toxic or pyrophoric service needs a helium leak test and a safety review, and the spark test and heaters apply only to some products.
  • Quantity, lead time, discount and total — the quantity range, the minimum lead time, the discount that needs approval or is blocked, and the stated total, which must match the total computed from the quoted items.

Any gas and chemical delivery subsystem quote that touches these rules is in scope — not only the scenarios pictured here. Not covered: other product lines, services and build-to-drawing details. The product categories, cleanroom classes, materials and tests are named on the public Ultra Clean Holdings pages; the material split, channel limits, service rules, lead-time and discount thresholds and every price are modeled from comparable products and are not published by the company, so all figures are illustrative.

The AI sales agentBuilds the subsystem 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.

6 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 material that must suit the productOnly with care
Stainless steel fits gas panels and manifolds and plastics fit chemical modules, so every material needs a rule that looks at the product.
The product and the material are in one request and checked together on every call.
Limits that depend on the productOnly with care
Each product has its own channel maximum, so the rule needs a separate list by product.
The product and the channel count are in one request, so the limit is checked on every call.
Service that needs a particular material or testOnly with workarounds
Corrosive and toxic service each add requirements on material and leak test, and toxic service also needs a person to review.
The service, the material and the test are all in the request, so every dependency is decided together.
Options that apply to some products onlyOnly with workarounds
Heaters, the spark test and the Class 100 build are not offered on every product, so each option needs its own product check.
The option and the product are in the request and checked on every call.
A discount with an approval band and a block limitOnly with workarounds
The two thresholds must be tested together so the right one fires.
The discount is compared with both thresholds on every call and returns a block, an approval or nothing.
A stated total that must match the computed totalOnly with workarounds
The total is built from several items, so a separate rule must recompute it and compare it with the stated figure.
The stated total is a field of the request and is compared with the computed total on every call, together with every other finding.

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.

141 channels is outside the range of 1 to 40 for this product.
2The material chosen does not suit the product; gas panels and valve manifolds are stainless steel and chemical delivery modules are all-plastic.
3Corrosive gas service needs electropolished 316L stainless steel.
4Toxic or pyrophoric gas service needs a safety engineering review.
5Toxic or pyrophoric service needs a helium leak test.
6The electrical permeability (spark) test applies to plastic assemblies only.
7Heaters are quoted on gas panels and manifolds only.
8Chemical delivery modules are assembled in Class 1,000 or ISO 7 cleanrooms.
9A lead time of 5 weeks is below the 6 week minimum.
10A lead time under 10 weeks on more than 12 channels needs planning to confirm capacity.
1151 units is outside the quoted range of 1 to 50.
12A discount of 9 percent is above 8 percent and needs sales management approval.
13A discount of 16 percent is above the 15 percent maximum.
14The quote states a total of 11,899,999 cents; the modeled prices give 11,900,000 cents.

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. Product, material and build: A chemical delivery module in stainless steel with heaters, a Class 100 build, a spark test, a short lead time and a 12 percent discount

The agent's request. Please quote 2 chemical delivery modules in standard 316L stainless steel, for inert gas service, assembled in a Class 100 cleanroom, with a spark test, heaters, a lead time of 4 weeks, 6 channels and a 12 percent discount, for a total of 10,229,200 cents. It is ready to send.

Verdict: Blocked.

  • Material does not suit the product. The material chosen does not suit the product; gas panels and valve manifolds are stainless steel and chemical delivery modules are all-plastic.
  • Spark test is for plastic only. The electrical permeability (spark) test applies to plastic assemblies only.
  • Heaters are for gas products. Heaters are quoted on gas panels and manifolds only.
  • Class 100 assembly not for chemical modules. Chemical delivery modules are assembled in Class 1,000 or ISO 7 cleanrooms.
  • Lead time under 6 weeks. A lead time of 4 weeks is below the 6 week minimum.
  • Stated total does not match. The quote states a total of 10,229,200 cents; the modeled prices give 10,279,200 cents.
  • Discount needs approval. A discount of 12 percent is above 8 percent and needs sales management approval.

2. Product, material and build: The agent resubmits after fixing the errors

The agent's request. Please quote 2 chemical delivery modules in PFA, for inert gas service, assembled in a Class 1000 cleanroom, with a spark test, a lead time of 14 weeks, 6 channels and a 5 percent discount, for a total of 11,806,200 cents. It is ready to send.

Verdict: Allowed.

  • Price summary. 4,898,000 cents per unit, 11,806,200 cents total including engineering at modeled prices.

3. Toxic service and channel count: A valve manifold for toxic gas service with no leak test, 44 channels and a 10 percent discount

The agent's request. Please quote 2 valve manifolds in electropolished 316L stainless steel, for toxic or pyrophoric gas service, assembled in a Class 100 cleanroom, with no leak test, a lead time of 16 weeks, 44 channels and a 10 percent discount, for a total of 22,570,000 cents. It is ready to send.

Verdict: Blocked.

  • Channel count outside the product range. 44 channels is outside the range of 1 to 40 for this product.
  • Toxic service needs a helium leak test. Toxic or pyrophoric service needs a helium leak test.
  • Discount needs approval. A discount of 10 percent is above 8 percent and needs sales management approval.
  • Toxic service needs safety review. Toxic or pyrophoric gas service needs a safety engineering review.

4. Toxic service and channel count: The agent resubmits after fixing the errors

The agent's request. Please quote 2 valve manifolds in electropolished 316L stainless steel, for toxic or pyrophoric gas service, assembled in a Class 100 cleanroom, with a helium leak test, a lead time of 16 weeks, 40 channels and a 10 percent discount, for a total of 21,220,000 cents. It is ready to send.

Verdict: Needs approval.

  • Discount needs approval. A discount of 10 percent is above 8 percent and needs sales management approval.
  • Toxic service needs safety review. Toxic or pyrophoric gas service needs a safety engineering review.
  • Price summary. 10,400,000 cents per unit, 21,220,000 cents total including engineering at modeled prices.

5. Corrosive service and test: A gas panel for corrosive service in standard steel with a spark test and 30 channels

The agent's request. Please quote 2 process tool gas panels in standard 316L stainless steel, for corrosive gas service, assembled in a Class 100 cleanroom, with a spark test, a lead time of 16 weeks, 30 channels, for a total of 25,780,000 cents. It is ready to send.

Verdict: Blocked.

  • Channel count outside the product range. 30 channels is outside the range of 1 to 24 for this product.
  • Corrosive service needs electropolished steel. Corrosive gas service needs electropolished 316L stainless steel.
  • Spark test is for plastic only. The electrical permeability (spark) test applies to plastic assemblies only.

6. Corrosive service and test: The agent resubmits after fixing the errors

The agent's request. Please quote 2 process tool gas panels in electropolished 316L stainless steel, for corrosive gas service, assembled in a Class 100 cleanroom, with a helium leak test, a lead time of 16 weeks, 24 channels, for a total of 25,900,000 cents. It is ready to send.

Verdict: Allowed.

  • Price summary. 11,700,000 cents per unit, 25,900,000 cents total including engineering at modeled 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 a quote for a gas and chemical delivery subsystem: a process tool gas panel, a valve manifold or an all-plastic chemical delivery module. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the product rules 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 2 chemical delivery modules in standard 316L stainless steel, for inert gas service, assembled in a Class 100 cleanroom. It is ready to send. The rest of the request is on screen.
  4. Before it goes out, the agent calls the validator.
  5. Blocked, with six findings and one approval, all in one answer. The material chosen does not suit the product; gas panels and valve manifolds are stainless steel and chemical delivery modules are all-plastic. The electrical permeability (spark) test applies to plastic assemblies only. Heaters are quoted on gas panels and manifolds only. The other three are on screen.
  6. The validator returned six 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 quote are checked against each other and against the stated total.
  7. What this shows: for a material that must suit the product, a rules-based CPQ is error-prone for this.
  8. Now the agent resubmits.
  9. The corrected request. It changes the material to PFA, the cleanroom class to Class 1000, the heaters to no, the lead time to 14 weeks, the discount to 5 percent and the total stated on the quote to 11,806,200 cents.
  10. It calls the validator again.
  11. This time the validator clears it.
  12. All seven earlier findings are cleared.
  13. Seven 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. Now you can try the other scenarios yourself.