Predictika · Communications API committed-spend quotes

An AI sales agent's committed-spend quote, checked before it goes out

An AI sales agent drafts a committed-spend contract for a communications API: the plan, the product, the monthly volume, the term, the annual commit and any extra discount. Before the quote goes out, the agent sends it to Predictika, which checks it against a declarative model of plan levels, volume minimums, terms, the commit against estimated usage and discount authority, compares the net total the agent states with the one it computes, and reports every problem it finds in one pass.

Scope: Twilio style committed-spend contract quotes: plan, product, monthly volume, term, annual commit and extra discount. See everything this validator covers

This validator models Twilio style committed-spend contract quotes for communications APIs: plan, product, monthly volume, term, annual commit and extra discount. It handles any contract made of these plans, products and terms, not just the scenarios shown below.

  • Plans and products — Growth, Scale and Enterprise plans with Programmable Messaging SMS, Email API, Verify, Programmable Voice and WhatsApp Business Messaging, each with the plan level it needs, a minimum monthly volume and, for Voice, a minimum term.
  • Annual commit — A minimum commit for each plan, and the commit compared with the estimated annual usage spend: a block below 60 percent, a deal desk review below 80 percent and a review above 130 percent.
  • Usage estimate — Estimated list usage less volume, plan and term discounts and any extra discount, in illustrative US dollars.
  • Term and extra discount — 1, 2 or 3 year terms, extra discount authority by plan (8, 12 and 18 percent) with approval above 5 percent, and a sales director review for contracts above USD 2,000,000.
  • The stated net total — A check that the net total the agent states matches the contract value the validator computes, which is the annual commit times the term.

Any committed-spend quote that touches these rules is in scope — not only the scenarios pictured here. Not covered: other Twilio products, overage billing after signing, phone number, carrier surcharge and support fees, real list prices and partner program terms, or multi-currency quoting.

The AI sales agentBuilds the committed-spend 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.

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
Product eligibility by plan level, monthly volume and termOnly with care
Dependency rules across three fields; they work, but each new product needs its own rows.		Limits are catalog data; the checks follow from the choices.
Volume, plan and term discounts building an estimated usage spendEasy
Lookups on volume, plan and term.		The same calculation, driven from the catalog.
A commit compared with a computed usage estimate, with a block floor and two review bandsOnly with workarounds
The estimate has to be computed before the commit can be compared with it, and three thresholds need three rules kept in order.		The estimate is computed from the choices and compared with all three thresholds on every call.
Extra discount authority by plan, with an approval pathOnly with care
Layered plan bands and approval thresholds; reliable only while every band is kept in step.		Bands and thresholds are data; findings arrive together.
The total the agent statesOnly with workarounds
Comparing a stated total with a computed one means a separate rule per quote layout, and price rules rarely run on stated figures.		The stated net total is a field in the request and is checked against the computed total.
Every problem returned in one answerOnly with workarounds
Rules tend to fire as the quote is saved, 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.

    
1An annual commit of USD 110,000 is below 80 percent of the estimated annual usage spend of USD 156,988.80; deal desk must approve the overage exposure.
2An annual commit of USD 250,000 is above 130 percent of the estimated annual usage spend of USD 156,988.80; deal desk must confirm the customer will use it.
3Contract value of USD 6,000,000 is above USD 2,000,000 and needs a sales director review.
410 percent extra discount is above the 5 percent a rep can give; approvable up to 12 percent for this plan.
5An annual commit of USD 9,000 is below the minimum of USD 10,000 for the Growth Plan.
6Programmable Voice is not available on the Growth Plan; it needs a higher plan.
7Programmable Voice needs a term of at least 2 years; 1 were quoted.
8Verify needs at least 2 thousand units a month; 1 were quoted.
9A term of 4 years is not offered; use 1, 2 or 3 years.
1020 percent extra discount exceeds the 12 percent limit for this plan.
11An annual commit of USD 100,000 is below 60 percent of the estimated annual usage spend of USD 235,483.20.
12The quoted net total of USD 135,000 does not match the USD 150,000 the validator computed.

  

  

    
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. Annual commit: The total is right; the commit is too small for the usage

    
The agent's request. This is for a retailer sending about 3 million order and delivery alerts by text message a month. Twilio Scale Plan with Programmable Messaging SMS for 3 million messages a month, one year, with an annual commit of USD 100,000. Net total USD 100,000. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Commit Below Usage Estimate. An annual commit of USD 100,000 is below 60 percent of the estimated annual usage spend of USD 235,483.20.
    • 
    

    
2. Annual commit: The agent resubmits after fixing the errors

    
The agent's request. This is for a retailer sending about 3 million order and delivery alerts by text message a month. Twilio Scale Plan with Programmable Messaging SMS for 3 million messages a month, one year, with an annual commit of USD 200,000. Net total USD 200,000. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Price summary. Estimated price: USD 200,000 contract value over 1 years at an annual commit of USD 200,000. Estimated annual usage spend USD 235,483.20 from USD 284,400 list after 10 percent volume, 8 percent plan, 0 percent term and 0 percent extra discount; discount limit 12 percent, for 3000 thousand units a month.
    • 
    

    
3. Voice term: Everything fits except the term

    
The agent's request. This is for a support-center operator moving its outbound calling to a cloud voice platform. Twilio Scale Plan with Programmable Voice for 1 million voice minutes a month, one year, with an annual commit of USD 140,000. Net total USD 140,000. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Product Term. Programmable Voice needs a term of at least 2 years; 1 were quoted.
    • 
    

    
4. Voice term: The agent resubmits after fixing the errors

    
The agent's request. This is for a support-center operator moving its outbound calling to a cloud voice platform. Twilio Scale Plan with Programmable Voice for 1 million voice minutes a month, two years, with an annual commit of USD 140,000 a year. Net total USD 280,000. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Price summary. Estimated price: USD 280,000 contract value over 2 years at an annual commit of USD 140,000. Estimated annual usage spend USD 142,427.04 from USD 168,000 list after 5 percent volume, 8 percent plan, 3 percent term and 0 percent extra discount; discount limit 12 percent, for 1000 thousand units a month.
    • 
    

    
5. The total the agent states: The agent's own total is checked too

    
The agent's request. This is for a fintech sending about 2 million one-time passcodes by text message a month. Twilio Scale Plan with Programmable Messaging SMS for 2 million messages a month, one year, with an annual commit of USD 150,000. Net total USD 135,000. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Quoted Net Total. The quoted net total of USD 135,000 does not match the USD 150,000 the validator computed.
    • 
    

    
6. The total the agent states: The agent resubmits after fixing the errors

    
The agent's request. This is for a fintech sending about 2 million one-time passcodes by text message a month. Twilio Scale Plan with Programmable Messaging SMS for 2 million messages a month, one year, with an annual commit of USD 150,000. Net total USD 150,000. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Price summary. Estimated price: USD 150,000 contract value over 1 years at an annual commit of USD 150,000. Estimated annual usage spend USD 156,988.80 from USD 189,600 list after 10 percent volume, 8 percent plan, 0 percent term and 0 percent extra discount; discount limit 12 percent, for 2000 thousand units a month.
    • 
    

    
7. Review by a person: Valid, but not final without a person

    
The agent's request. This is for a start-up planning 2 million text messages a month that wants to keep its commitment low. Twilio Scale Plan with Programmable Messaging SMS for 2 million messages a month, one year, with an annual commit of USD 110,000. Net total USD 110,000. It is ready to send.

    
Verdict: Needs approval.

    
    
      
  • Commit Review. An annual commit of USD 110,000 is below 80 percent of the estimated annual usage spend of USD 156,988.80; deal desk must approve the overage exposure.
    • 
      
  • Price summary. Estimated price: USD 110,000 contract value over 1 years at an annual commit of USD 110,000. Estimated annual usage spend USD 156,988.80 from USD 189,600 list after 10 percent volume, 8 percent plan, 0 percent term and 0 percent extra discount; discount limit 12 percent, for 2000 thousand units a month.
    • 
    

  

  

    
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 drafts a committed-spend contract for a communications API: the plan, the product, the monthly volume, and more. Before the quote goes out, the agent sends it to Predictika, which checks it against a declarative model of plan levels, volume minimums, terms, the commit against estimated usage and discount authority, compares the net total the agent states with the one it computes, and reports every problem it finds in one pass.
    2. 
      
  2. I will start with the first request by the agent.
    3. 
      
  3. The agent says. This is for a retailer sending about 3 million order and delivery alerts by text message a month. Twilio Scale Plan with Programmable Messaging SMS for 3 million messages a month, one year, with an annual commit of USD 100,000. The rest of the request is on screen. Net total USD 100,000.
    4. 
      
  4. Before it goes out, the agent calls the validator.
    5. 
      
  5. Blocked, with one finding, all in one answer. An annual commit of USD 100,000 is below 60 percent of the estimated annual usage spend of USD 235,483.20.
    6. 
      
  6. The validator returned one blocking finding 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.
    7. 
      
  7. What this shows: for extra discount authority by plan, with an approval path, a rules-based CPQ is error-prone for this.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the annual commit to $200,000 and the net total stated on the quote to $200,000.
    10. 
      
  10. It calls the validator again.
    11. 
      
  11. This time the validator clears it.
    12. 
      
  12. The validator also returns the line-item quote, on screen.
    13. 
      
  13. The earlier finding is cleared, and the validator's total, USD 200,000, matches the quote.
    14. 
      
  14. One problem caught in one call, fixed in one pass, and a clean $200,000 quote. That is the value of Predictika's validator: catching mistakes before they reach the customer. Now you can try the other scenarios yourself.
    15.