Predictika · Everpure FlashArray deals

An AI sales agent's FlashArray deal, checked before it goes out

An AI sales agent prepares an Everpure FlashArray deal for a customer: array model, raw capacity, Evergreen program, site power and rack space, discount. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the sizing, site-fit and Evergreen program rules and returns a verdict, every reason and the capacity figures it computed.

Scope: Everpure FlashArray//X, //XL, //C and //E deal quotes, Evergreen program terms and discount authority. See everything this validator covers

This validator models Everpure FlashArray//X, //XL, //C and //E deal quotes and the Evergreen programs sold with them, not just the scenarios shown below.

  • Systems — FlashArray//X (X20, X50, X70, X90), //XL (XL130, XL170, XL190), //C (C50, C70, C90) and //E, with each model's largest datasheet configuration: maximum raw capacity, peak power and rack units.
  • Sizing — effective capacity from raw capacity at the 5 to 1 datasheet data reduction and at the ratio the quote claims, checked against the capacity the customer needs and against the capacity the quote states.
  • Site fit — peak power draw against the site's power, and rack space against the site's rack units.
  • Evergreen programs — Evergreen//Forever (controller upgrade eligibility, flat renewal pricing, Capacity Consolidation, Ever Agile), //Foundation, //Flex (reserve commitment) and //One (service tiers and their minimum reserves, energy and paid power SLAs, rate lock, renewal cap, opt-out notice, reserve changes, buffer, Unified Data Repository terms).
  • FlashArray//E — sold only under Evergreen//One as a Unified Data Repository subscription, from 750 TB raw, with no S3 object access.
  • Discount authority — who can approve a discount by deal size, and when a discount needs approval or is refused.

Any FlashArray//X, //XL, //C or //E deal that touches sizing, site fit, these Evergreen programs or discounts is in scope — not only the scenarios pictured here. Not covered: list prices and per-TiB rates (the deal value is an input, so this validator produces no price build-up or line-item quote), FlashBlade, Portworx and Pure Fusion, and Everpure's own CPQ platform and discount matrix. Assumptions: FlashArray//E effective capacity (3.7 to 1) is inferred and its power and rack space are not published; discount limits are placeholders.

The AI sales agentBuilds the FlashArray deal 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
Capacity that depends on model, raw capacity and data reductionOnly with care
Effective capacity comes from the model's ratio, the raw capacity and the claimed data reduction, so a separate rule has to recompute it for every model and compare it with the need and with the figure the quote states.
Model, raw capacity, claimed ratio, the customer's need and the stated capacity are fields of one request, and the validator computes the capacity itself.
A choice that fixes one hold and changes anotherOnly with workarounds
Moving to a larger model changes the raw capacity limit, the power draw, the rack space, the deal size and the discount authority, and the rules that watch them fire one at a time.
Every rule is checked on every call, so one change is seen against all the others.
Program terms that depend on the offer and the deal typeOnly with workarounds
Controller-upgrade timing, Capacity Consolidation caps, the return window and Ever Agile promotion limits each belong to one program and have to be written and kept apart for Forever, Foundation, Flex and One.
Offer, deal type, years since upgrade, credits and promotions are fields of one request and every program rule is checked together.
Service tiers and SLAs with their own minimumsOnly with workarounds
Each Evergreen//One tier has its own reserve minimum, energy and power SLA thresholds and renewal, opt-out and term limits, and an SLA with no published minimum needs a human decision.
Tier, reserve, SLAs, renewal rise, notice period and term are fields of one request, and an undocumented SLA is returned as an approval.
Every problem returned in one answerOnly with workarounds
Rules tend to fire as the quote 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 deal; approval findings need a person to sign off; everything else is computed.

1700 TB raw is above the 695 TB the FlashArray//X70 R5 supports in its largest configuration.
2The customer needs 2,000 TB effective but 300 TB raw on the FlashArray//X70 R5 gives only 1,079.13 TB even at the claimed 5.0 to 1 data reduction.
3The quote meets the 2,300 TB effective need only by claiming 6.0 to 1 data reduction; the datasheet basis is 5 to 1 (2,158.26 TB). Needs a sizing review.
4The FlashArray//X70 R5 draws up to 1,616 W at peak; the site allows 1,500 W.
5The FlashArray//X70 R5 can take up to 9 rack units with shelves; the site offers 8. Needs a site survey.
6FlashArray//E is block or unified block and file only; the customer needs S3 object or unified file and object.
7FlashArray//E is sold only under Evergreen//One as a Unified Data Repository subscription, not as Forever on tier No Evergreen//One tier.
8FlashArray//E starts at 750 TB raw; the quote has 600 TB.
9The Premium tier has a minimum reserve of 50 TiB; the quote reserves 40 TiB.
10Unified Data Repository is a 36-month subscription; the quote has 24 months.
11The paid power and rack SLA on the Performance tier needs a reserve of at least 200 TiB; the quote reserves 150 TiB.
12The paid power and rack SLA needs utilization of 80 percent or more and 200 to 240 V power; the site has 70 percent and 120 V.
13The requested SLA has no published minimum for the Premium tier. Product management must confirm eligibility.
14The buffer guarantee covers 25 percent of the reserve up to 1 PiB (100.0 TiB here); the quote asks 150 TiB.
15The existing rate holds on an expansion only up to 2,000 TiB in total and only above 80 percent utilization; this one reaches 2,500 TiB at 90 percent.
16An Evergreen//One renewal can rise by at most 20 percent over the prior rate; the quote raises it 25 percent.
17Opting out of auto-renewal needs 60 days notice; the request gives 30 days.
18Evergreen//Forever renewal pricing stays flat to the initial reseller rate; the quote is 8 percent higher while annual PPI inflation is 3 percent (7 or less).
19Annual PPI inflation of 9 percent is above 7, so flat renewal pricing is open to negotiation; the quote is 8 percent higher.
20A controller upgrade on the FlashArray//X70 R5 needs 3 consecutive years since the last upgrade; the customer has 2.
21Evergreen//Foundation does not include Ever Modern controller upgrades, Capacity Consolidation or Ever Agile.
22Capacity Consolidation credit is capped at 50 percent of the new capacity and 25 percent of the subscription; the quote asks 60 and 0.
23Replaced capacity must come back within 45 days unless a non-return SKU is bought; the quote allows 60 days.
24Ever Agile trade-in credit cannot be combined with other promotions.
25Evergreen//Flex needs a reserve commitment of at least 30 percent of effective capacity; the quote has 20 percent.
26An Evergreen//One reserve commitment can only increase; the quote lowers it from 400 to 300 TiB.
2712 percent discount is above the 5 percent a rep can give; approvable up to 15 percent on this deal ($400,000).
2820 percent discount exceeds the 15 percent anyone can approve on a deal of $400,000.
2920 percent discount is above the 5 percent a rep can give; approvable up to 25 percent on this deal ($1,500,000).
30A controller upgrade on the FlashArray//X90 R5 needs 3 consecutive years since the last upgrade; the customer has 2.
31The customer needs 4,000 TB effective but 1,000 TB raw on the FlashArray//X90 R5 gives only 3,645.4 TB even at the claimed 5.0 to 1 data reduction.
32The quote states 4,000.0 TB effective capacity; the FlashArray//X90 R5 gives 3,645.4 TB at the claimed 5.0 to 1 data reduction.
3330 percent discount exceeds the 25 percent anyone can approve on a deal of $1,200,000.
3420 percent discount is above the 5 percent a rep can give; approvable up to 25 percent on this deal ($1,200,000).

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. Model, program terms and discount: An X70 controller upgrade at two years with 1,500 TB raw, Ever Agile with another promotion, a Capacity Consolidation credit over the cap, a stated capacity that does not match and a 20 percent discount

The agent's request. Please quote a FlashArray//X70 controller upgrade on Evergreen//Forever, 1,500 TB raw for a customer who needs 6,000 TB effective, 2 years since the last upgrade, with Ever Agile and another promotion, a 60 percent Capacity Consolidation credit with the old capacity back in 60 days, a 20 percent discount on a deal of USD 400,000, stating 6,000 TB effective capacity. It is ready to send.

Verdict: Blocked.

  • Raw capacity above the model maximum. 1,500 TB raw is above the 695 TB the FlashArray//X70 R5 supports in its largest configuration.
  • Too small for the capacity the customer needs. The customer needs 6,000 TB effective but 1,500 TB raw on the FlashArray//X70 R5 gives only 5,395.65 TB even at the claimed 5.0 to 1 data reduction.
  • Controller upgrade eligibility. A controller upgrade on the FlashArray//X70 R5 needs 3 consecutive years since the last upgrade; the customer has 2.
  • Capacity Consolidation credit cap. Capacity Consolidation credit is capped at 50 percent of the new capacity and 25 percent of the subscription; the quote asks 60 and 0.
  • Replaced capacity return window. Replaced capacity must come back within 45 days unless a non-return SKU is bought; the quote allows 60 days.
  • Ever Agile cannot stack with promotions. Ever Agile trade-in credit cannot be combined with other promotions.
  • Stated capacity does not match. The quote states 6,000.0 TB effective capacity; the FlashArray//X70 R5 gives 5,395.65 TB at the claimed 5.0 to 1 data reduction.
  • Discount above anyone's authority. 20 percent discount exceeds the 15 percent anyone can approve on a deal of $400,000.

2. Model, program terms and discount: The agent resubmits after fixing the errors

The agent's request. Please quote a FlashArray//XL170 new purchase on Evergreen//Forever, 1,650 TB raw for a customer who needs 6,000 TB effective, with Ever Agile, a 50 percent Capacity Consolidation credit with the old capacity back in 45 days, a 5 percent discount on a deal of USD 1,200,000, stating 6,197.89 TB effective capacity. It is ready to send.

Verdict: Allowed.

  • Capacity summary. 1,650 TB raw on the FlashArray//XL170 R5 gives 6,197.89 TB effective at the 5 to 1 datasheet basis and 6,197.89 TB at the claimed 5.0 to 1.

3. Evergreen//One renewal: An Evergreen//One renewal on Premium with a 40 TiB reserve, an energy SLA, a 25 percent rise and 30 days notice

The agent's request. Please quote a FlashArray//X70 renewal on Evergreen//One, Premium tier with a 40 TiB reserve, with the energy efficiency SLA, a renewal 25 percent above the prior rate, 30 days opt-out notice. It is ready to send.

Verdict: Blocked.

  • Tier minimum reserve not met. The Premium tier has a minimum reserve of 50 TiB; the quote reserves 40 TiB.
  • Renewal rise above 20 percent. An Evergreen//One renewal can rise by at most 20 percent over the prior rate; the quote raises it 25 percent.
  • Opt-out notice under 60 days. Opting out of auto-renewal needs 60 days notice; the request gives 30 days.
  • SLA with no published minimum. The requested SLA has no published minimum for the Premium tier. Product management must confirm eligibility.

4. Evergreen//One renewal: The agent resubmits after fixing the errors

The agent's request. Please quote a FlashArray//X70 renewal on Evergreen//One, Performance tier with a 350 TiB reserve, with the energy efficiency SLA, a renewal 18 percent above the prior rate, 60 days opt-out notice. It is ready to send.

Verdict: Allowed.

5. FlashArray//E: FlashArray//E at 600 TB raw under Evergreen//Forever with S3 object access

The agent's request. Please quote a FlashArray//E new purchase on Evergreen//Forever, 600 TB raw, with S3 object access. It is ready to send.

Verdict: Blocked.

  • FlashArray//E has no S3 object access. FlashArray//E is block or unified block and file only; the customer needs S3 object or unified file and object.
  • FlashArray//E starts at 750 TB raw. FlashArray//E starts at 750 TB raw; the quote has 600 TB.
  • FlashArray//E is sold only under Evergreen//One UDR. FlashArray//E is sold only under Evergreen//One as a Unified Data Repository subscription, not as Forever on tier No Evergreen//One tier.

6. FlashArray//E: The agent resubmits after fixing the errors

The agent's request. Please quote a FlashArray//E new purchase on Evergreen//One, Unified Data Repository tier with a 800 TiB reserve, 800 TB raw, a 36-month term. It is ready to send.

Verdict: Allowed.

  • Capacity summary. 800 TB raw on the FlashArray//E gives 2,960.0 TB effective at the 5 to 1 datasheet basis and 2,960.0 TB at the claimed 5.0 to 1.

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 Everpure FlashArray deal for a customer: array model, raw capacity, Evergreen program, and more. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the sizing, site-fit and Evergreen program rules and returns a verdict, every reason and the capacity figures it computed.
  2. I will start with the first request by the agent.
  3. The agent says. Please quote a FlashArray//X70 controller upgrade on Evergreen//Forever, 1,500 TB raw for a customer who needs 6,000 TB effective, 2 years since the last upgrade. It is ready to send. The rest of the request is on screen.
  4. Before it goes out, the agent calls the validator.
  5. Blocked, with eight findings, all in one answer. 1,500 TB raw is above the 695 TB the FlashArray//X70 R5 supports in its largest configuration. The customer needs 6,000 TB effective but 1,500 TB raw on the FlashArray//X70 R5 gives only 5,395.65 TB even at the claimed 5.0 to 1 data reduction. A controller upgrade on the FlashArray//X70 R5 needs 3 consecutive years since the last upgrade; the customer has 2. The other five are on screen.
  6. The validator returned eight 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 deal are checked against each other.
  7. What this shows: for capacity that depends on model, raw capacity and data reduction, a rules-based CPQ is error-prone for this.
  8. Now the agent resubmits.
  9. The corrected request. It changes the array to FlashArray//XL170, the deal type to New Purchase, the raw capacity to 1,650 TB, the discount to 5 percent, the deal value to $1,200,000 and the effective capacity stated on the quote to 6,197.89 TB.
  10. It calls the validator again.
  11. This time the validator clears it.
  12. All eight earlier findings are cleared.
  13. Eight problems caught in one call, fixed in one pass, and a clean deal. That is the value of Predictika's validator: catching mistakes before the deal is signed.