Predictika · Foodservice distribution orders

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

An AI sales agent prepares an order of Sysco-brand products for a foodservice customer, with a tier, a truck, a delivery day and a discount. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of brand, tier, storage, delivery and discount rules and returns a verdict, every reason and the total it computed.

Scope: Orders of seven Sysco brands (Wholesome Farms, Buckhead Pride, Block & Barrel, Portico, Arrezzio, Baker's Source and Casa Solana) in the Premium, Classic and Reliance tiers, with truck type, shelf-life cover, delivery day, rush delivery, cut-off, case quantity and discount. See everything this validator covers

This validator models Sysco-brand foodservice orders, not just the scenarios shown below.

  • Brand and tier — which of the Premium, Classic and Reliance tiers is offered for each brand, and the case price that follows from the brand and tier.
  • Storage and shelf life — dry, refrigerated and frozen brands, the truck each needs, and how many cases the customer can use before the shelf life ends at its weekly usage.
  • Delivery — the delivery minimum, the small-order fee, rush delivery (not available on weekends) and the order cut-off.
  • Discount, large orders and stated total — discounts that need a sales manager or are over the limit, orders of 200 cases or more that the warehouse confirms, and the order total the quote states against the total computed.

Any Sysco-brand foodservice order that touches these rules is in scope — not only the scenarios pictured here. Not covered: other Sysco brands and products, non-Sysco products, equipment and supplies, and customer-specific contract pricing.

The AI sales agentBuilds the foodservice 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.

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 brand that comes only in some tiersOnly with care
Each brand needs its own list of offered tiers, and the list has to be kept in step with the price table.		The brand and the tier are fields of one request and are checked together on every call.
Chilled and frozen products on a dry-only truckOnly with care
The storage type of each brand has to be tied to the truck type with a separate rule for each brand.		The brand, its storage type and the truck are in one request and checked together on every call.
Quantity against the shelf-life coverOnly with workarounds
The cover depends on the storage type and the weekly usage, so a derived limit has to be computed before the quantity can be compared with it.		The cover is computed from the weekly usage and compared with the quantity on every call.
Minimum, small-order fee and weekend rush togetherOnly with care
Delivery minimum, fee and rush rules interact, and each one is usually a separate rule that fires at a different step.		Subtotal, fees, delivery day and rush are in one request and are checked together.
A stated total that must match the computed totalOnly with workarounds
The total is built from several lines, so a separate rule is needed to 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.
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.

    
1The selected Sysco tier is not offered for the selected brand.
2This brand ships refrigerated or frozen and needs a multi-temperature truck; the route has a dry-only truck.
331 cases are more than the customer can use before the shelf life ends (about 30 cases at the current weekly usage).
4121 cases are more than the customer can use before the shelf life ends (about 120 cases at the current weekly usage).
5The order subtotal of 380 dollars is below the 500 dollar delivery minimum.
6Rush delivery is not available for weekend delivery.
7A discount of 8 percent is above 5 percent; a sales manager should approve it.
8A discount of 15 percent is above 5 percent; a sales manager should approve it.
9A discount of 18 percent is above the 15 percent limit.
10The order missed the cut-off time; the delivery date needs to be confirmed with the operating company.
11An order of 200 cases should be confirmed with the warehouse.
12The quote states a total of 100 cents; the modeled prices give 78,500 cents.
1340 cases are more than the customer can use before the shelf life ends (about 15 cases at the current weekly usage).
14A discount of 20 percent is above the 15 percent limit.
15The order has 0 cases; at least 1 is needed.
16The order subtotal of 0 dollars is below the 500 dollar delivery minimum.

  

  

    
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. Tier, truck, shelf life and discount: A hotel order of Buckhead Pride with the wrong tier, truck, quantity and discount

    
The agent's request. Please quote 40 cases of Buckhead Pride in the Reliance tier for a hotel, delivered on a weekday, with a 20 percent discount, for a total of $5,024.00. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Tier not offered for this brand. The selected Sysco tier is not offered for the selected brand.
    • 
      
  • Chilled or frozen product on a dry-only truck. This brand ships refrigerated or frozen and needs a multi-temperature truck; the route has a dry-only truck.
    • 
      
  • More stock than can be used in time. 40 cases are more than the customer can use before the shelf life ends (about 36 cases at the current weekly usage).
    • 
      
  • Discount above 15 percent. A discount of 20 percent is above the 15 percent limit.
    • 
    

    
2. Tier, truck, shelf life and discount: The agent resubmits after fixing the errors

    
The agent's request. Please quote 36 cases of Buckhead Pride in the Premium tier for a hotel, delivered on a weekday, with a 5 percent discount, for a total of $6,498.00. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Order summary. 36 cases at 190 dollars each, delivery fees 0 dollars, modeled total of 649,800 cents.
    • 
    

    
3. Delivery minimum, truck and weekend rush: A weekend rush delivery of frozen Portico cases

    
The agent's request. Please quote 3 cases of Portico in the Classic tier for a restaurant, delivered on a weekend with rush delivery, for a total of $520.00. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Chilled or frozen product on a dry-only truck. This brand ships refrigerated or frozen and needs a multi-temperature truck; the route has a dry-only truck.
    • 
      
  • Below the delivery minimum. The order subtotal of 420 dollars is below the 500 dollar delivery minimum.
    • 
      
  • No weekend rush delivery. Rush delivery is not available for weekend delivery.
    • 
    

    
4. Delivery minimum, truck and weekend rush: The agent resubmits after fixing the errors

    
The agent's request. Please quote 5 cases of Portico in the Classic tier for a restaurant, delivered on a weekday with rush delivery, for a total of $800.00. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Order summary. 5 cases at 140 dollars each, delivery fees 100 dollars, modeled total of 80,000 cents.
    • 
    

    
5. Large order, discount and stated total: A large school order of Casa Solana with a discount over the limit and a total that does not match

    
The agent's request. Please quote 240 cases of Casa Solana in the Classic tier for a school, delivered on a weekday, with an 18 percent discount, placed after the order cut-off, for a total of $8,568.00. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Discount above 15 percent. A discount of 18 percent is above the 15 percent limit.
    • 
      
  • Stated total does not match. The quote states a total of 857,640 cents; the modeled prices give 826,560 cents.
    • 
      
  • Order cut-off missed. The order missed the cut-off time; the delivery date needs to be confirmed with the operating company.
    • 
      
  • Large order to confirm. An order of 240 cases should be confirmed with the warehouse.
    • 
    

    
6. Large order, discount and stated total: The agent resubmits after fixing the errors

    
The agent's request. Please quote 240 cases of Casa Solana in the Classic tier for a school, delivered on a weekday, with an 8 percent discount, placed after the order cut-off, for a total of $9,273.60. It is ready to send.

    
Verdict: Needs approval.

    
    
      
  • Discount above 5 percent. A discount of 8 percent is above 5 percent; a sales manager should approve it.
    • 
      
  • Order cut-off missed. The order missed the cut-off time; the delivery date needs to be confirmed with the operating company.
    • 
      
  • Large order to confirm. An order of 240 cases should be confirmed with the warehouse.
    • 
      
  • Order summary. 240 cases at 42 dollars each, delivery fees 0 dollars, modeled total of 927,360 cents.
    • 
    

  

  

    
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 of Sysco-brand products for a foodservice customer, with a tier, a truck, a delivery day and a discount. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of brand, tier, storage, delivery and discount rules and returns a verdict, every reason and the total it computed.
    2. 
      
  2. I will start with the first request by the agent.
    3. 
      
  3. The agent says. Please quote 40 cases of Buckhead Pride in the Reliance tier for a hotel, delivered on a weekday, with a 20 percent discount, for a total of $5,024.00. It is ready to send.
    4. 
      
  4. Before it goes out, the agent calls the validator.
    5. 
      
  5. Blocked, with four findings, all in one answer. The selected Sysco tier is not offered for the selected brand. This brand ships refrigerated or frozen and needs a multi-temperature truck; the route has a dry-only truck. 40 cases are more than the customer can use before the shelf life ends (about 36 cases at the current weekly usage). The other one is on screen.
    6. 
      
  6. The validator returned four 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 chilled and frozen products on a dry-only truck, a rules-based CPQ is error-prone for this.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the tier to premium, the cases to 36, the truck to multi-temperature, the discount to 5 percent and the total stated on the quote to 6,498.00.
    10. 
      
  10. It calls the validator again.
    11. 
      
  11. This time the validator clears it.
    12. 
      
  12. All four earlier findings are cleared.
    13. 
      
  13. Four problems caught in one call, fixed in one pass, 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.
    14.