Predictika · Managed security service quotes

An AI sales agent's managed security service quote, checked before it goes out

An AI sales agent prepares a quote for a managed Microsoft Sentinel security service: a contract term, a Microsoft agreement type, daily log volume, log sources, retention, incident response hours, add-ons, onboarding details and a requested discount. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of this service and returns a verdict, every reason and the monthly fee it computed.

Scope: Managed Sentinel Service quotes: onboarding prerequisites, contract term and Microsoft agreement, log volume and sources, retention and long-term storage, incident response hours, add-ons, exclusions, discount and the stated monthly fee. See everything this validator covers

This validator models Managed Sentinel Service quotes, not just the scenarios shown below.

  • Onboarding and exclusions — two authorized contacts, an empty subscription, tenant details, an owner login, the Azure Lighthouse offer, no personal or payment card data in logs, and the work the service excludes: project and design work, the wider Azure estate and vulnerability management.
  • Retention, hours and add-ons — 90 days of retention with Azure Data Explorer for longer, 20 incident response hours included and extra hours bought, and the Azure Support Service for Enterprise Agreement and Customer Agreement customers.
  • Size, discount and stated fee — ingestion bands and extra log sources, large estates and custom content sent for confirmation, a requested discount above the term discount sent to a person and over 15 percent blocked, and the stated monthly fee compared with the computed fee.

Any Managed Sentinel quote that touches these rules is in scope — not only the scenarios pictured here. Not covered: Softcat's other services and the resale of vendor licenses. Public versus modeled: the service components, the 90-day retention, the 20 included incident response hours, the onboarding prerequisites and the exclusions come from Softcat's public service definition; the size bands, term discounts, thresholds and every price (pence, in pounds sterling before VAT) are modeled by Predictika from comparable products, are illustrative and are not published by Softcat.

The AI sales agentBuilds the managed service 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
Prerequisites that must all be in place before onboardingOnly with care
Contacts, subscription, tenant details, owner login and Lighthouse acceptance are each a separate rule that has to be kept in step with the service definition.		All onboarding details are fields of one quote and are checked together on every call.
Retention that needs a different storageOnly with care
Retention over 90 days depends on whether long-term storage is quoted, so the rule has to look at both.		Retention and storage are fields of one quote and are compared on every call.
Add-ons that depend on the customer's agreementOnly with care
The Azure Support Service is for two agreement types only, so the add-on rule has to look at the agreement.		Add-on and agreement are fields of one quote and are checked together on every call.
A stated monthly fee that must match the computed feeOnly with workarounds
The fee is built from the size band, extra sources, add-ons and the term discount, so a separate rule is needed to recompute it and compare it with the stated figure.		The stated fee is a field of the request and is compared with the computed monthly fee on every call.
Large estates and discounts that go to a personOnly with care
A separate approval path with its own thresholds that must not overlap the blocking rules.		Blocks and approvals come back together; an approval-only quote is not blocked.

  

  

    
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.

    
191 days of retention is wanted but Log Analytics keeps 90 days; add Azure Data Explorer long-term storage.
2The service needs at least 2 authorised customer contacts; the quote has 1.
3The onboarding prerequisites are incomplete: the tenant hostname, tenant ID, subscription ID and Azure region have not been provided.
4The Azure subscription for the Sentinel deployment must be empty.
5An owner-level login for the subscription is required for deployment.
6The customer must accept the Azure Lighthouse offer and app registrations so analysts and engineers can be given access.
7Logs must not contain personally identifiable information.
8Logs must not contain payment card information; storing or managing card data is excluded from the service.
930 incident response hours are needed but the service includes 20 and the quote buys 0 more.
10The Softcat Azure Support Service (SD50) is for customers with a Microsoft Enterprise Agreement or Customer Agreement.
11Support for the wider Azure environment is excluded; only the Sentinel service and its deployed resources are covered.
12Project and design work, including professional services, is outside the managed service and needs a separate engagement.
13Vendor patch notification and vulnerability management are excluded from the service.
14At least 1 log source must be onboarded; the quote has 0.
15150 GB a day is above the largest standard size band; a person should confirm sizing.
16High-complexity custom log work (API sources, parsers, connectors, automations or enrichment) may be charged separately; a person should scope and price it.
1761 log sources is a large onboarding; a person should confirm the onboarding plan.
18800 days of retention needs a long-term storage plan agreed before deployment.
19Softcat asks for 30 days notice of additions to the supported products; the quote gives 14.
20The requested discount of 10 percent is above the 6 percent term discount; a person should approve it.
21A requested extra discount of 20 percent is above the 15 percent limit of this model.
22The quote states a monthly fee of 507,599 pence; the modeled price gives 507,600 pence.
23The quote states a monthly fee of 540,000 pence; the modeled price gives 507,600 pence.
24365 days of retention is wanted but Log Analytics keeps 90 days; add Azure Data Explorer long-term storage.
25120 GB a day is above the largest standard size band; a person should confirm sizing.
26Softcat asks for 30 days notice of additions to the supported products; the quote gives 10.

  

  

    
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. Managed Sentinel onboarding: A Managed Sentinel Service quote with 400 days of retention, 40 incident response hours, project work and an unfinished onboarding

    
The agent's request. Please quote the Managed Sentinel Service for 36 months on a Microsoft Enterprise Agreement, with 20 GB of logs a day from 8 log sources, 400 days of retention, 40 incident response hours needed and 1 authorized contact, and also project work, for a stated monthly fee of 507,600 pence. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Fewer than two authorised contacts. The service needs at least 2 authorised customer contacts; the quote has 1.
    • 
      
  • Subscription is not empty. The Azure subscription for the Sentinel deployment must be empty.
    • 
      
  • Azure Lighthouse offer not accepted. The customer must accept the Azure Lighthouse offer and app registrations so analysts and engineers can be given access.
    • 
      
  • Personal data in logs. Logs must not contain personally identifiable information.
    • 
      
  • Retention over 90 days without long-term storage. 400 days of retention is wanted but Log Analytics keeps 90 days; add Azure Data Explorer long-term storage.
    • 
      
  • Incident response hours short. 40 incident response hours are needed but the service includes 20 and the quote buys 0 more.
    • 
      
  • Project or design work requested. Project and design work, including professional services, is outside the managed service and needs a separate engagement.
    • 
    

    
2. Managed Sentinel onboarding: The agent resubmits after fixing the errors

    
The agent's request. Please quote the Managed Sentinel Service for 36 months on a Microsoft Enterprise Agreement, with 20 GB of logs a day from 8 log sources, 400 days of retention, 40 incident response hours needed and 2 authorized contacts, plus Azure Data Explorer storage and 20 extra incident response hours, for a stated monthly fee of 549,900 pence. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Fee summary. Monthly fee of 549,900 pence after a 6 percent term discount, 3,300 pounds of one-off charges, 20,126,400 pence over the contract (illustrative modeled prices in GBP, before VAT).
    • 
    

    
3. Azure Support Service add-on: A Managed Sentinel Service quote on a Cloud Solution Provider agreement with the Azure Support Service and a 20 percent discount

    
The agent's request. Please quote the Managed Sentinel Service for 36 months on a Cloud Solution Provider agreement, with 20 GB of logs a day from 8 log sources, 90 days of retention, 20 incident response hours needed and 2 authorized contacts, plus the Azure Support Service, and also support for the wider Azure estate, with a requested discount of 20 percent, for a stated monthly fee of 568,700 pence. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Payment card data in logs. Logs must not contain payment card information; storing or managing card data is excluded from the service.
    • 
      
  • Azure Support Service needs EA or MCA. The Softcat Azure Support Service (SD50) is for customers with a Microsoft Enterprise Agreement or Customer Agreement.
    • 
      
  • Wider Azure estate requested. Support for the wider Azure environment is excluded; only the Sentinel service and its deployed resources are covered.
    • 
      
  • Requested discount over 15 percent. A requested extra discount of 20 percent is above the 15 percent limit of this model.
    • 
    

    
4. Azure Support Service add-on: The agent resubmits after fixing the errors

    
The agent's request. Please quote the Managed Sentinel Service for 36 months on a Microsoft Enterprise Agreement, with 20 GB of logs a day from 8 log sources, 90 days of retention, 20 incident response hours needed and 2 authorized contacts, plus the Azure Support Service, with a requested discount of 10 percent, for a stated monthly fee of 568,700 pence. It is ready to send.

    
Verdict: Needs approval.

    
    
      
  • Extra discount above the term discount. The requested discount of 10 percent is above the 6 percent term discount; a person should approve it.
    • 
      
  • Fee summary. Monthly fee of 568,700 pence after a 6 percent term discount, 0 pounds of one-off charges, 20,473,200 pence over the contract (illustrative modeled prices in GBP, before VAT).
    • 
    

    
5. Large Sentinel estate: A 60-month quote for 120 GB of logs a day from 70 sources, with vulnerability management and a stated monthly fee

    
The agent's request. Please quote the Managed Sentinel Service for 60 months on a Microsoft Enterprise Agreement, with 120 GB of logs a day from 70 log sources, 90 days of retention, 20 incident response hours needed and 2 authorized contacts, and also vulnerability management, for a stated monthly fee of 2,446,000 pence. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Vulnerability management requested. Vendor patch notification and vulnerability management are excluded from the service.
    • 
      
  • Stated monthly fee does not match. The quote states a monthly fee of 2,446,000 pence; the modeled price gives 2,466,000 pence.
    • 
      
  • Ingestion above 100 GB a day. 120 GB a day is above the largest standard size band; a person should confirm sizing.
    • 
      
  • More than 60 log sources. 70 log sources is a large onboarding; a person should confirm the onboarding plan.
    • 
    

    
6. Large Sentinel estate: The agent resubmits after fixing the errors

    
The agent's request. Please quote the Managed Sentinel Service for 60 months on a Microsoft Enterprise Agreement, with 120 GB of logs a day from 70 log sources, 90 days of retention, 20 incident response hours needed and 2 authorized contacts, for a stated monthly fee of 2,466,000 pence. It is ready to send.

    
Verdict: Needs approval.

    
    
      
  • Ingestion above 100 GB a day. 120 GB a day is above the largest standard size band; a person should confirm sizing.
    • 
      
  • More than 60 log sources. 70 log sources is a large onboarding; a person should confirm the onboarding plan.
    • 
      
  • Fee summary. Monthly fee of 2,466,000 pence after a 10 percent term discount, 0 pounds of one-off charges, 147,960,000 pence over the contract (illustrative modeled prices in GBP, before VAT).
    • 
    

  

  

    
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 managed Microsoft Sentinel security service: a contract term, a Microsoft agreement type, daily log volume, and more. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of this service and returns a verdict, every reason and the monthly fee it computed.
    2. 
      
  2. I will start with the first request by the agent.
    3. 
      
  3. The agent says. Please quote the Managed Sentinel Service for 36 months on a Microsoft Enterprise Agreement, with 20 GB of logs a day from 8 log sources. It is ready to send. The rest of the request is on screen.
    4. 
      
  4. Before it goes out, the agent calls the validator.
    5. 
      
  5. Blocked, with seven findings, all in one answer. The service needs at least 2 authorised customer contacts; the quote has 1. The Azure subscription for the Sentinel deployment must be empty. The customer must accept the Azure Lighthouse offer and app registrations so analysts and engineers can be given access. The other four are on screen.
    6. 
      
  6. The validator returned seven 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.
    7. 
      
  7. What this shows: for retention that needs a different storage, a rules-based CPQ is error-prone for this.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the extra incident response hours bought to 20, the authorized contacts to 2, the empty subscription to yes, the lighthouse offer accepted to yes and the monthly fee as stated on the quote to 549,900. Two more changes are on screen.
    10. 
      
  10. It calls the validator again.
    11. 
      
  11. This time the validator clears it.
    12. 
      
  12. All seven earlier findings are cleared.
    13. 
      
  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.
    14.