Predictika · Gas engine orders for generator sets

An AI sales agent's PSI gas engine order, checked before it goes out

An AI sales agent prepares an order for PSI gas engines that drive generator sets: the engine model, fuel, speed, duty rating, number of engines and redundancy against a site load. Before it 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 figures it computed.

Scope: PSI gas engines 2.4LNA, 4X, 6.7LNA, 6.7LT, 8.1LT and 8.8LNA on natural gas or LPG at 1500 or 1800 rpm, with standby and prime ratings. See everything this validator covers

This validator models orders for six PSI gas engines, not just the scenarios shown below. The model names, fuels and the kWe ratings at 1500 and 1800 rpm follow public PSI spec sheets; the sizing logic and approval rules are modeled from comparable products and are not PSI's published rules.

  • Models, fuel and speed — the 2.4LNA, 4X (4.3 L), 6.7LNA, 6.7LT, 8.1LT and 8.8LNA, on natural gas or propane (LPG) at 1500 or 1800 rpm; bi-fuel or dual-fuel operation is listed only for the 2.4LNA, 4X and 8.8LNA.
  • Duty ratings — the standby (ESP) and prime (PRP) generator output in kWe for each model, fuel and speed; the 2.4LNA sheet lists only the standby rating.
  • Sizing — the engines online (the order less any redundant units) must cover the site load; the total installed kWe the quote states is compared with the quantity times the rating.
  • Approvals — a bi-fuel or dual-fuel configuration, or a site above 1,000 m, needs a person to confirm (needs approval).

Any PSI gas engine order that touches these rules is in scope — not only the scenarios pictured here. Not covered: prices (PSI publishes no price list for these engines, so this validator produces no price build-up or line-item quote), the other PSI engine families, continuous (COP) ratings, emissions certification, and generator set accessories. The sizing and approval rules are modeled from comparable products, not company-published.

The AI sales agentBuilds the engine order 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
Ratings that differ by model, fuel, speed and dutyOnly with workarounds
Forty-odd combinations of rating figures have to be kept as lookup tables or long rule lists, and every sheet revision means editing them.
The rating is looked up from the model, fuel, speed and duty on every call, and the figure is shown.
Engines online after redundant units against the site loadOnly with care
A computed capacity compared with a customer figure; it works once a rule is written for each combination.
The capacity is computed and compared with the site load every time, with both numbers shown.
A duty rating that the sheet does not publishOnly with care
A restriction per model and duty; one more rule for each exception.
The missing rating is detected from the lookup and reported by name.
A stated installed total that must match quantity times ratingOnly with care
A stated figure is a free-entry field; a rule can compare it with a computed value.
The stated figure travels with the order and is compared with the computed one.
Fuel options listed for only some modelsOnly with care
A product rule per fuel and model.
The fuel and model are in one request and are checked together.
Every problem returned in one answerOnly with workarounds
Rules tend to fire as 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 quote; approval findings need a person to sign off; everything else is computed.

1No rating is published for this model at the chosen duty (the spec sheet lists only the standby rating).
2The site needs 55 kWe; the units online give 54 kWe.
3Bi-fuel or dual-fuel operation is not offered on this model; it is listed for natural gas and propane only.
4Bi-fuel or dual-fuel operation is requested; the sheet gives no separate rating, so a person should confirm the fuel system and rating with PSI.
5The site needs 83 kWe; the units online give 82 kWe.
61 redundant unit(s) leave nothing online in an order of 1 engine(s).
7The site needs 150 kWe; the units online give 0 kWe.
8The site is at 1,001 m; ratings are subject to site conditions, so a person should confirm any derating with PSI.
9The quote states 200 kWe installed in total; the quoted engines give 240 kWe.
10The site needs 97 kWe; the units online give 96 kWe.
11The site needs 150 kWe; the units online give 100 kWe.
12The quote states 200 kWe installed in total; the quoted engines give 150 kWe.
13The site is at 1,500 m; ratings are subject to site conditions, so a person should confirm any derating with PSI.
14The site is at 1,200 m; ratings are subject to site conditions, so a person should confirm any derating with PSI.

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. Cold storage site, 260 kWe: Three 6.7LT engines on natural gas, one redundant, for a 260 kWe load

The agent's request. Please quote 3 PSI 6.7LT gas engines on natural gas at 1800 rpm, rated for prime (PRP) duty, with 1 of them redundant, for a site load of 260 kWe at 300 m. The quote states 400 kWe installed in total. It is ready to send.

Verdict: Blocked.

  • Online capacity below the site load. The site needs 260 kWe; the units online give 202 kWe.
  • Stated installed total does not match. The quote states 400 kWe installed in total; the quoted engines give 303 kWe.

2. Cold storage site, 260 kWe: The agent resubmits after fixing the errors

The agent's request. Please quote 3 PSI 8.1LT gas engines on natural gas at 1800 rpm, rated for prime (PRP) duty, with 1 of them redundant, for a site load of 260 kWe at 300 m. The quote states 396 kWe installed in total. It is ready to send.

Verdict: Allowed.

  • Order summary. 3 8.1LT engine(s), 132 kWe each at the chosen duty, 396 kWe installed, 264 kWe available with 2 online.

3. Small clinic, 25 kWe: Two 2.4LNA engines rated for prime duty

The agent's request. Please quote 2 PSI 2.4LNA gas engines on natural gas at 1500 rpm, rated for prime (PRP) duty, with 1 of them redundant, for a site load of 25 kWe at 100 m. It is ready to send.

Verdict: Blocked.

  • Rating not published for this duty. No rating is published for this model at the chosen duty (the spec sheet lists only the standby rating).

4. Small clinic, 25 kWe: The agent resubmits after fixing the errors

The agent's request. Please quote 2 PSI 4X gas engines on bi-fuel at 1500 rpm, rated for prime (PRP) duty, with 1 of them redundant, for a site load of 25 kWe at 100 m. It is ready to send.

Verdict: Needs approval.

  • Bi-fuel or dual-fuel configuration. Bi-fuel or dual-fuel operation is requested; the sheet gives no separate rating, so a person should confirm the fuel system and rating with PSI.
  • Order summary. 2 4X engine(s), 38 kWe each at the chosen duty, 76 kWe installed, 38 kWe available with 1 online.

5. Small clinic, 25 kWe: The agent resubmits after fixing the errors

The agent's request. Please quote 2 PSI 4X gas engines on natural gas at 1500 rpm, rated for prime (PRP) duty, with 1 of them redundant, for a site load of 25 kWe at 100 m. It is ready to send.

Verdict: Allowed.

  • Order summary. 2 4X engine(s), 38 kWe each at the chosen duty, 76 kWe installed, 38 kWe available with 1 online.

6. Water plant, 150 kWe: Two 8.8LNA engines with one redundant and a stated total

The agent's request. Please quote 2 PSI 8.8LNA gas engines on natural gas at 1800 rpm, rated for prime (PRP) duty, with 1 of them redundant, for a site load of 150 kWe at 400 m. The quote states 200 kWe installed in total. It is ready to send.

Verdict: Blocked.

  • Online capacity below the site load. The site needs 150 kWe; the units online give 95 kWe.
  • Stated installed total does not match. The quote states 200 kWe installed in total; the quoted engines give 190 kWe.

7. Water plant, 150 kWe: The agent resubmits after fixing the errors

The agent's request. Please quote 3 PSI 8.8LNA gas engines on natural gas at 1800 rpm, rated for prime (PRP) duty, with 1 of them redundant, for a site load of 150 kWe at 400 m. The quote states 285 kWe installed in total. It is ready to send.

Verdict: Allowed.

  • Order summary. 3 8.8LNA engine(s), 95 kWe each at the chosen duty, 285 kWe installed, 190 kWe available with 2 online.

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 PSI gas engines that drive generator sets: the engine model, fuel, speed, and more. Before it 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 figures it computed.
  2. I will start with the first request by the agent.
  3. The agent says. Please quote 3 PSI 6.7LT gas engines on natural gas at 1800 rpm, rated for prime (PRP) duty, with 1 of them redundant. The quote states 400 kWe installed in total. The rest of the request is on screen.
  4. Before it goes out, the agent calls the validator.
  5. Blocked, with two findings, all in one answer. The site needs 260 kWe; the units online give 202 kWe. The quote states 400 kWe installed in total; the quoted engines give 303 kWe.
  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 quote are checked against each other and against the stated total.
  7. What this shows: for a stated installed total that must match quantity times rating, a rules-based CPQ is error-prone for this.
  8. Now the agent resubmits.
  9. The corrected request. It changes the engine model to 8.1LT and the stated installed total to 396 kWe.
  10. It calls the validator again.
  11. This time the validator clears it.
  12. Both earlier findings are cleared, and the validator's figure matches the quote.
  13. Two 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.