Predictika · Everpure Evergreen//One catalog

An AI sales agent's Evergreen//One quote, checked before it goes out

An AI sales agent prepares an Evergreen//One quote for a customer beyond FlashArray: a file, object, block or AI tier, the reserve, performance and term, and sometimes the ransomware recovery SLA. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the published tier minimums, fixed terms, protocol fit and SLA eligibility, and returns a verdict and every reason.

Scope: Everpure Evergreen//One catalog quotes beyond FlashArray: block and file, Unified Fast File and Object, Unified Data Repository and AI tiers, and the ransomware recovery SLA. See everything this validator covers

This validator models the Evergreen//One service catalog on the tiers below and handles any quote involving them — not just the scenarios shown on this page.

Placeholders and inferences: the discount limits (5, 15 and 25 percent) are Predictika placeholders because Everpure does not publish its discount matrix, and the AI tier serving file and object is inferred, not stated in Everpure's documents. The tier minimums and terms come from Everpure's public documents; no prices are used.

  • Block and file tiers — Ultra, Premium, Performance, Adaptive and Capacity, with their minimum reserves and the Adaptive performance reserve.
  • Unified Fast File and Object tiers — Premium, Performance, Standard and Adaptive, with minimum reserves and the Adaptive performance reserve.
  • Unified Data Repository and AI tiers — Unified Data Repository (file and object, and block and file) with its minimum reserve and fixed 36-month term; the AI tier with its 30 GB/s performance reserve and 12-month term.
  • Protocol fit — Whether the tier serves block or S3 object, checked against what the customer needs.
  • Ransomware recovery SLA — Eligibility (an add-on for Evergreen//One customers) and the combinations Everpure's public documents do not cover, which go to a person.
  • Stated figures and discount authority — The tier minimums the quote states are compared with the catalog, and who can approve a discount by deal size.

Any Evergreen//One quote on these tiers is in scope, not only the scenarios pictured here. Not covered: prices and per-TiB rates (the deal value is an input, so this validator produces no price build-up or line-item quote), the other Evergreen//One SLAs beyond the ransomware recovery add-on, and Evergreen//Forever, //Foundation and //Flex (see the FlashArray validator). Assumptions: the AI tier is assumed to serve file and object, which is inferred; the discount limits (5, 15 and 25 percent) are placeholders.

The AI sales agentBuilds the Evergreen//One 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
Adaptive tier quoted with too little throughputOnly with care
Capacity passes. The GB/s minimum is a second threshold that differs by tier, so each tier needs its own validation rule, and the performance reserve is easy to leave blank.
Blocks on the GB/s minimum, 5 for Adaptive file and object and 1 for Adaptive block and file.
Unified Data Repository quoted at a short termOnly with care
The term is a free-entry field. A rule per tier can compare it with the fixed 36 months, but it must be written and kept for every fixed-term tier.
Blocks; the fixed term is stored on the tier record and compared on every call.
S3 object on a block and file tierOnly with care
Protocol support has to be recorded per tier and checked against a customer need that is rarely a field on the quote.
Blocks on protocol fit; the need travels with the request.
Ransomware recovery SLA added for a customer who is not on Evergreen//OneOnly with workarounds
Eligibility belongs to the customer account, not the quote. A rule can check it only if the program is copied onto the quote as a field.
Blocks on eligibility.
Tier minimums stated in the quote that differ from the catalogOnly with workarounds
A validation rule can compare a stated figure with a catalog value only if the stated figure is a field on the quote, which it rarely is.
The stated minimums travel with the quote and are compared with the catalog on every call.
Combination the public documents do not cover (SLA on the AI tier)Only with workarounds
A CPQ allows or rejects. It has no way to say that nothing published covers the combination and a person must decide.
Routes to a person for approval and says why.

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.

1the Adaptive (Unified Fast File and Object) tier needs a performance reserve of at least 5 GB/s; the quote has 3 GB/s.
2the quote states a minimum performance reserve of 3 GB/s; the Adaptive (Unified Fast File and Object) tier minimum is 5 GB/s.
3the Unified Data Repository (Block and File) tier is a 36-month term; the quote has 24 months.
4the Premium (Unified Block and File) tier does not serve S3 object; the customer needs it.
5the ransomware recovery SLA is an add-on for Evergreen//One customers; this customer is not on Evergreen//One.
6availability of the ransomware recovery SLA on the AI tier is not published; product management must confirm.
7the Performance (Unified Block and File) tier has a minimum of 50 TiB; the quote reserves 40 TiB.
8the quote states a minimum reserve of 100 TiB; the Capacity (Unified Block and File) tier minimum is 200 TiB.
9the AI tier needs a performance reserve of at least 30 GB/s; the quote has 20 GB/s.
1030 percent discount exceeds the 25 percent anyone can approve on a deal of $1,500,000.
1120 percent discount is above the 5 percent a rep can give; approvable up to 25 percent on this deal ($1,500,000).
12the Performance (Unified Fast File and Object) tier does not serve block; the customer needs it.

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. Adaptive file and object tier: Three wrong figures on an Adaptive file and object quote

The agent's request. Please quote Orchard Media on the Evergreen//One Adaptive Unified Fast File and Object tier: a 100 TiB reserve with a 3 GB/s performance reserve, at a deal value of $300,000. The tier minimums are 50 TiB of capacity and 3 GB/s of performance, which this order meets. It is ready to send.

Verdict: Blocked.

  • Performance reserve below the tier minimum. The Adaptive (Unified Fast File and Object) tier needs a performance reserve of at least 5 GB/s; the quote has 3 GB/s.
  • Stated capacity minimum does not match. The quote states a minimum reserve of 50 TiB; the Adaptive (Unified Fast File and Object) tier minimum is 100 TiB.
  • Stated performance minimum does not match. The quote states a minimum performance reserve of 3 GB/s; the Adaptive (Unified Fast File and Object) tier minimum is 5 GB/s.

2. Adaptive file and object tier: The agent resubmits after fixing the errors

The agent's request. Please quote Orchard Media on the Evergreen//One Adaptive Unified Fast File and Object tier: a 100 TiB reserve with a 5 GB/s performance reserve, at a deal value of $300,000. The tier minimums are 100 TiB of capacity and 5 GB/s of performance, which this order meets. It is ready to send.

Verdict: Allowed.

  • All rules satisfied. Every rule checked is satisfied.

3. Unified Data Repository term: 24-month term on a fixed 36-month tier

The agent's request. Please quote Kestrel Bank on the Evergreen//One Unified Data Repository (Block and File) tier: a 500 TiB reserve on a 24-month term, at a deal value of $900,000. The tier minimum is 500 TiB of capacity, with no performance minimum. It is ready to send.

Verdict: Blocked.

  • Term does not match the tier. The Unified Data Repository (Block and File) tier is a 36-month term; the quote has 24 months.

4. Unified Data Repository term: The agent resubmits after fixing the errors

The agent's request. Please quote Kestrel Bank on the Evergreen//One Unified Data Repository (Block and File) tier: a 500 TiB reserve on a 36-month term, at a deal value of $900,000. The tier minimum is 500 TiB of capacity, with no performance minimum. It is ready to send.

Verdict: Allowed.

  • All rules satisfied. Every rule checked is satisfied.

5. Ransomware recovery SLA: SLA for a customer not on Evergreen//One

The agent's request. Please add the ransomware recovery SLA to a Performance Unified Block and File quote for Meridian Health: a 100 TiB reserve at a deal value of $300,000. Meridian Health is an Evergreen//Forever customer. The tier minimum is 50 TiB of capacity, with no performance minimum. It is ready to send.

Verdict: Blocked.

  • Ransomware recovery SLA not available to this customer. The ransomware recovery SLA is an add-on for Evergreen//One customers; this customer is not on Evergreen//One.

6. Ransomware recovery SLA: The agent resubmits after fixing the errors

The agent's request. Please add the ransomware recovery SLA to a Performance Unified Block and File quote for Meridian Health: a 100 TiB reserve at a deal value of $300,000. Meridian Health is an Evergreen//One customer. The tier minimum is 50 TiB of capacity, with no performance minimum. It is ready to send.

Verdict: Allowed.

  • All rules satisfied. Every rule checked is satisfied.

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 Evergreen//One quote for a customer beyond FlashArray: a file, object, block or AI tier, and more. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the published tier minimums, fixed terms, protocol fit and SLA eligibility, and returns a verdict and every reason.
  2. I will start with the first request by the agent.
  3. The agent says. Please quote Orchard Media on the Evergreen//One Adaptive Unified Fast File and Object tier: a 100 TiB reserve with a 3 GB/s performance reserve, at a deal value of $300,000. The tier minimums are 50 TiB of capacity and 3 GB/s of performance, which this order meets. The rest of the request is on screen.
  4. Before it goes out, the agent calls the validator.
  5. Blocked, with three findings, all in one answer. The Adaptive (Unified Fast File and Object) tier needs a performance reserve of at least 5 GB/s; the quote has 3 GB/s. The quote states a minimum reserve of 50 TiB; the Adaptive (Unified Fast File and Object) tier minimum is 100 TiB. The quote states a minimum performance reserve of 3 GB/s; the Adaptive (Unified Fast File and Object) tier minimum is 5 GB/s.
  6. The validator returned three 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. What this shows: for tier minimums stated in the quote that differ from the catalog, a rules-based CPQ can do this only with workarounds.
  8. Now the agent resubmits.
  9. The corrected request. It changes the performance reserve to 5 GB/s, the minimum reserve the quote states to 100 TiB and the minimum performance the quote states to 5 GB/s.
  10. It calls the validator again.
  11. This time the validator clears it.
  12. All three earlier findings are cleared, and the validator's figure matches the quote.
  13. Three 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.