Predictika · Data center colocation orders

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

An AI sales agent prepares a colocation order for Iron Mountain Data Centers: cabinets, a cage, a suite or a modular data center (MOD) in one of eleven North American markets, with power per cabinet, redundancy, cross connects, Cloud On-Ramp, Metro Wave, the contract term and a discount. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of how those choices fit together and returns a verdict, every reason and the monthly total it computed.

Scope: One Iron Mountain Data Centers colocation order: offering, market, power and redundancy, cross connects, Cloud On-Ramp, Metro Wave, FedRAMP Ready, term, discount and the monthly total. See everything this validator covers

This validator models Iron Mountain Data Centers colocation orders, not just the scenarios shown below.

  • Offering and size — colocation cabinets, data center cages (6 or more cabinets, 30 kW or more), private suites and MODs, with the number of cabinets checked against what each offering is for.
  • Market — the eleven markets, with MODs only in strategic markets and Metro Wave and Cloud On-Ramp only where they are modeled as available.
  • Power and redundancy — kilowatts per cabinet against the modeled limit, high-density power that engineering must confirm, and N+1 against 2N.
  • FedRAMP Ready — the FedRAMP Ready environment is checked against the one data center that lists it, Northern Virginia VA-1.
  • Cross connects — the number per cabinet against the modeled limit, with approval above eight per cabinet.
  • Term and discount — the discount against the limit for the 12, 36 or 60 month term, with approval at 10 percent or more.
  • Stated total — the monthly total the order states against the total computed from the modeled prices.

Any Iron Mountain colocation order that touches these rules is in scope — not only the scenarios pictured here. Not covered: records management, shredding and digital solutions. Every price here is modeled and illustrative, not published by Iron Mountain.

The AI sales agentBuilds the colocation order from the conversation with the customer 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
A cage that needs enough cabinets and powerOnly with care
A cage is for 6 or more cabinets and 30 kW or more, and a suite for 20 or more, so each offering needs its own quantity and power rules that read other fields.
The offering, the cabinets and the kilowatts are fields of one request and are checked together on every call.
Options that exist only in some marketsOnly with care
FedRAMP Ready, MODs, Metro Wave and Cloud On-Ramp each depend on the market, so every option needs a rule that reads the market, and the list grows with every market added.
The market and every option are in one request and are checked against each other on every call.
A discount limit that depends on the termOnly with workarounds
The allowed discount changes with the 12, 36 or 60 month term, so a separate lookup has to sit beside every discount rule.
The term and the discount are fields of the request and are checked against each other on every call.
Power per cabinet with an engineering thresholdOnly with care
One rule blocks power above the limit and another sends high-density power to engineering, so the two ranges have to be kept in step.
Both ranges are checked in the same pass, and blocks and approvals come back together.
A stated total that must match the computed totalOnly with workarounds
The total is built from power, space, fees, cross connects and options, 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 monthly 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 quote; approval findings need a person to sign off; everything else is computed.

1The order has 0 cabinets; at least 1 is needed.
24 cross connects is above the 12 per cabinet modeled limit for 0 cabinets.
36 individual colocation cabinets is modeled as too many; 6 or more cabinets belong in a cage.
4A data center cage is for 6 or more cabinets; the order has 5.
5A data center cage is for 30 or more kilowatts; the order totals 24 kW.
6A private data center suite is modeled for 20 or more cabinets; the order has 19.
7Modular data centers (MODs) are offered in strategic markets only; this market is modeled as not one of them.
8A+B 2N power is modeled as not available on MODs, which run with block N+1.
921 kW per cabinet is above the 20 kW per cabinet this model supports.
1020 kW per cabinet is high density; data center engineering should confirm cooling and power for the site.
11The FedRAMP Ready environment is listed for the Northern Virginia VA-1 data center; this market is modeled as not having one.
12Metro Wave is modeled as available in Northern Virginia, Phoenix, Chicago and New Jersey only.
13Cloud On-Ramp is modeled as not available in Ohio, Western Pennsylvania and Kansas City.
1440 cross connects for 4 cabinets is above 8 per cabinet; network engineering should confirm tray and panel space.
15The discount is above the limit modeled for this contract term (12 months 5 percent, 36 months 10 percent, 60 months 15 percent).
16A discount of 10 percent or more needs regional sales director approval.
17The order states a monthly total of 9,999 dollars; the modeled prices give 6,480 dollars.

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. Cabinets, discount and term: Six individual cabinets with a 15 percent discount on a 12-month term

The agent's request. Please quote 6 individual colocation cabinets in Northern Virginia at 8 kW per cabinet, with 2N redundancy and 12 cross connects, plus a 10 Gbps Metro Wave, on a 12-month term with a 15 percent discount. The monthly total is 15,487 dollars. It is ready to send.

Verdict: Blocked.

  • Too many cabinets for the cabinet offering. 6 individual colocation cabinets is modeled as too many; 6 or more cabinets belong in a cage.
  • Discount above the term limit. The discount is above the limit modeled for this contract term (12 months 5 percent, 36 months 10 percent, 60 months 15 percent).

2. Cabinets, discount and term: The agent resubmits after fixing the errors

The agent's request. Please quote a data center cage of 6 cabinets in Northern Virginia at 8 kW per cabinet, with 2N redundancy and 12 cross connects, plus a 10 Gbps Metro Wave, on a 12-month term with a 5 percent discount. The monthly total is 18,867 dollars. It is ready to send.

Verdict: Allowed.

  • Order summary. 6 cabinets at 8 kW each (48 kW), a monthly total of 18,867 dollars after discount.

3. Power per cabinet: A twelve-cabinet cage at 24 kW per cabinet

The agent's request. Please quote a data center cage of 12 cabinets in Northern Virginia at 24 kW per cabinet, with 2N redundancy and 12 cross connects, plus a 10 Gbps Metro Wave, on a 12-month term with a 5 percent discount. The monthly total is 65,835 dollars. It is ready to send.

Verdict: Blocked.

  • Power per cabinet above 20 kW. 24 kW per cabinet is above the 20 kW per cabinet this model supports.

4. Power per cabinet: The agent resubmits after fixing the errors

The agent's request. Please quote a data center cage of 12 cabinets in Northern Virginia at 12 kW per cabinet, with 2N redundancy and 12 cross connects, plus a 10 Gbps Metro Wave, on a 12-month term with a 5 percent discount. The monthly total is 38,475 dollars. It is ready to send.

Verdict: Needs approval.

  • High-density power above 10 kW per cabinet. 12 kW per cabinet is high density; data center engineering should confirm cooling and power for the site.
  • Order summary. 12 cabinets at 12 kW each (144 kW), a monthly total of 38,475 dollars after discount.

5. FedRAMP Ready and market: An eight-cabinet cage in Ohio that must be a FedRAMP Ready environment

The agent's request. Please quote a data center cage of 8 cabinets in Ohio at 6 kW per cabinet, with N+1 redundancy and 10 cross connects, plus a FedRAMP Ready environment, on a 36-month term with no discount. The monthly total is 13,280 dollars. It is ready to send.

Verdict: Blocked.

  • FedRAMP only at Northern Virginia VA-1. The FedRAMP Ready environment is listed for the Northern Virginia VA-1 data center; this market is modeled as not having one.

6. FedRAMP Ready and market: The agent resubmits after fixing the errors

The agent's request. Please quote a data center cage of 8 cabinets in Northern Virginia at 6 kW per cabinet, with N+1 redundancy and 10 cross connects, plus a FedRAMP Ready environment, on a 36-month term with no discount. The monthly total is 15,200 dollars. It is ready to send.

Verdict: Allowed.

  • Order summary. 8 cabinets at 6 kW each (48 kW), a monthly total of 15,200 dollars after discount.

7. Discount approval: A cage with a 10 percent discount on a 36-month term

The agent's request. Please quote a data center cage of 6 cabinets in Northern Virginia at 8 kW per cabinet, with 2N redundancy and 12 cross connects, plus a 10 Gbps Metro Wave, on a 36-month term with a 10 percent discount. The monthly total is 17,874 dollars. It is ready to send.

Verdict: Needs approval.

  • Discount of 10 percent or more. A discount of 10 percent or more needs regional sales director approval.
  • Order summary. 6 cabinets at 8 kW each (48 kW), a monthly total of 17,874 dollars after discount.

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 colocation order for Iron Mountain Data Centers: cabinets, a cage, a suite or a modular data center (MOD) in one of eleven North American markets. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of how those choices fit together and returns a verdict, every reason and the monthly total it computed.
  2. I will start with the first request by the agent.
  3. The agent says. Please quote 6 individual colocation cabinets in Northern Virginia at 8 kW per cabinet, with 2N redundancy and 12 cross connects. The monthly total is 15,487 dollars. The rest of the request is on screen.
  4. Before it goes out, the agent calls the validator.
  5. Blocked, with two findings, all in one answer. 6 individual colocation cabinets is modeled as too many; 6 or more cabinets belong in a cage. The discount is above the limit modeled for this contract term (12 months 5 percent, 36 months 10 percent, 60 months 15 percent).
  6. The validator returned two 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 a discount limit that depends on the term, a rules-based CPQ can do this only with workarounds.
  8. Now the agent resubmits.
  9. The corrected request. It changes the offering to data center cage, the discount to 5 percent and the monthly total stated on the order to 18,867 dollars.
  10. It calls the validator again.
  11. This time the validator clears it.
  12. Both earlier findings are cleared.
  13. Two 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.