Predictika · Agentic AI platform subscription quotes

An AI sales agent's TechM Orion subscription quote, checked before it goes out

An AI sales agent prepares a subscription quote for the TechM Orion agentic AI platform, with an edition, a deployment model, custom and library agents, governance checks, connectors, GPUs, a term and a discount. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the platform and its limits and returns a verdict, every reason and the total it computed.

Scope: TechM Orion platform subscriptions: three editions, cloud, on-premise and hybrid deployment, custom and library agents, governance checks, the VerifAI observer agent, connectors, fine-tuned models, GPUs, term and discount. See everything this validator covers

This validator models TechM Orion subscription quotes, not just the scenarios shown below.

  • Editions and agents — the Starter, Professional and Enterprise editions, the number of custom (Chat-to-Agent) and pre-built library agents each allows, and the governance checks per agent.
  • Deployment and options — cloud, on-premise and hybrid deployment and the GPUs they need, the VerifAI observer agent, the Email, SAP and Salesforce connectors and fine-tuned models, and which edition each option needs.
  • Term, discount and total — the subscription term, the discount and the approval it needs, early access, and the total the quote states, which must match the total computed from the prices.

Any TechM Orion subscription quote that touches these rules is in scope — not only the scenarios pictured here. Not covered: services and implementation, and other Tech Mahindra platforms. The platform name, the Chat-to-Agent builder, the library of pre-built agents, the governance checks, VerifAI, the connectors, fine-tuned models and the deployment options follow the public Tech Mahindra pages; the three editions, their limits, the compatibility rules, the term and discount rules and every price are modeled from comparable products and are not published by Tech Mahindra, so all figures are illustrative. The validator shows a computed total but no line-item table.

The AI sales agentBuilds the subscription 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
Edition limits on agents and governance checksOnly with care
Each edition has its own agent cap and governance cap, and the agent count is the sum of two fields, so each cap needs a calculated rule.		The editions, agent counts and governance checks are in one request and checked together on every call.
Options that depend on the edition, the deployment or the industryOnly with care
VerifAI for regulated industries, GPUs for on-premise and hybrid, and connectors and fine-tuning by edition each refer to a different choice, so the rule list grows with every option.		All the choices are in the request, so each dependency is checked on every call.
A discount and a term that each have a block limit and an approval bandOnly with workarounds
Each needs a block above one value and an approval above another, and the separate rules must agree with each other.		Each value is compared with its limits on every call and returns a block, an approval or nothing.
A stated total that must match the computed totalOnly with workarounds
The total is built from the platform, agents, options, deployment, GPUs, term and discount, so a separate rule must recompute it and compare it with the stated figure.		The stated total is a field of the request and is compared with the computed total on every call.
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 order; approval findings need a person to sign off; everything else is computed.

    
111 agents exceed the 10 agents this edition allows.
251 agents exceed the 50 agents this edition allows.
3501 agents exceed the 500 agents this edition allows.
4201 library agents requested; the pre-built library is quoted up to 200.
5The quote has 0 agents; at least 1 is needed.
631 governance checks per agent are outside the 30 per agent this edition allows.
711 governance checks per agent are outside the 10 per agent this edition allows.
861 governance checks per agent are outside the 60 per agent this edition allows.
9Banking and healthcare quotes need the VerifAI observer agent.
10The Starter edition is cloud-hosted only.
11On-premise and hybrid deployment need at least 4 NVIDIA GPUs; the quote has 3.
12On-premise and hybrid deployment need at least 4 NVIDIA GPUs; the quote has 0.
13A cloud deployment has no customer GPUs; the quote lists 2.
14The SAP and Salesforce enterprise adapters need the Professional edition or higher.
15Fine-tuned model support is part of the Enterprise edition only.
16A term of 11 months is outside the 12 to 60 months quoted.
17A term of 37 months needs a person to confirm price protection.
18A term of 60 months needs a person to confirm price protection.
19A term of 61 months is outside the 12 to 60 months quoted.
20A term of 61 months needs a person to confirm price protection.
21A discount of 11 percent needs sales leadership approval.
22A discount of 25 percent needs sales leadership approval.
23A discount of 26 percent is beyond what can be approved on a quote.
24Early access terms are set by the vendor team and need a person to confirm.
25A discount of 15 percent needs sales leadership approval.
26A term of 48 months needs a person to confirm price protection.
27The quote states a total of $305,999.00; the price list gives $306,000.00.
28The quote states a total of $306,000.01; the price list gives $306,000.00.
29The quote states a total of $0.07; the price list gives $306,000.00.
3016 agents exceed the 10 agents this edition allows.
31On-premise and hybrid deployment need at least 4 NVIDIA GPUs; the quote has 1.
32A term of 6 months is outside the 12 to 60 months quoted.
33A discount of 30 percent is beyond what can be approved on a quote.

  

  

    
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. Edition, deployment and options: A Starter edition quote for a bank with on-premise hosting, SAP, fine-tuned models and two GPUs

    
The agent's request. Please quote a Starter TechM Orion on-premise subscription for a banking and financial services customer with 3 custom agents, 5 library agents and 10 governance checks per agent, with fine-tuned models, the Email and SAP connectors, and 2 GPUs, on a 12 month term, with a discount of 0 percent, for a total of $202,800.00. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • VerifAI required in regulated industries. Banking and healthcare quotes need the VerifAI observer agent.
    • 
      
  • Starter is cloud only. The Starter edition is cloud-hosted only.
    • 
      
  • GPUs for on-premise and hybrid. On-premise and hybrid deployment need at least 4 NVIDIA GPUs; the quote has 2.
    • 
      
  • SAP and Salesforce connectors need Professional. The SAP and Salesforce enterprise adapters need the Professional edition or higher.
    • 
      
  • Fine-tuned models need Enterprise. Fine-tuned model support is part of the Enterprise edition only.
    • 
    

    
2. Edition, deployment and options: The agent resubmits after fixing the errors

    
The agent's request. Please quote a Enterprise TechM Orion hybrid subscription for a banking and financial services customer with 3 custom agents, 5 library agents and 10 governance checks per agent, with the VerifAI observer agent, fine-tuned models, the Email and SAP connectors, and 4 GPUs, on a 12 month term, with a discount of 0 percent, for a total of $638,400.00. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Order summary. 8 agent(s), 53,200 dollar(s) a month, 12 month term, estimated total of $638,400.00 at modeled prices.
    • 
    

    
3. Agents, governance and stated total: A Professional healthcare quote with many agents, customer GPUs and a stated total

    
The agent's request. Please quote a Professional TechM Orion cloud subscription for a healthcare customer with 45 custom agents, 10 library agents and 35 governance checks per agent, with the Email connector, and 2 GPUs, on a 12 month term, with a discount of 0 percent, for a total of $886,800.00. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Edition agent limit. 55 agents exceed the 50 agents this edition allows.
    • 
      
  • Governance checks per agent. 35 governance checks per agent are outside the 30 per agent this edition allows.
    • 
      
  • VerifAI required in regulated industries. Banking and healthcare quotes need the VerifAI observer agent.
    • 
      
  • GPUs on a cloud deployment. A cloud deployment has no customer GPUs; the quote lists 2.
    • 
      
  • Stated total does not match. The quote states a total of $886,800.00; the price list gives $888,000.00.
    • 
    

    
4. Agents, governance and stated total: The agent resubmits after fixing the errors

    
The agent's request. Please quote a Professional TechM Orion cloud subscription for a healthcare customer with 40 custom agents, 10 library agents and 30 governance checks per agent, with the VerifAI observer agent, the Email connector, on a 12 month term, with a discount of 0 percent, for a total of $834,000.00. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Order summary. 50 agent(s), 69,500 dollar(s) a month, 12 month term, estimated total of $834,000.00 at modeled prices.
    • 
    

    
5. Term, discount and early access: An Enterprise telecom quote with a 72 month term, a large discount and early access

    
The agent's request. Please quote a Enterprise TechM Orion cloud subscription for a telecom customer with 60 custom agents, 100 library agents and 40 governance checks per agent, with the Email connector, on a 72 month term in the early access program, with a discount of 30 percent, for a total of $7,560,000.00. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Subscription term. A term of 72 months is outside the 12 to 60 months quoted.
    • 
      
  • Discount beyond authority. A discount of 30 percent is beyond what can be approved on a quote.
    • 
      
  • Term over 36 months. A term of 72 months needs a person to confirm price protection.
    • 
      
  • Early access program. Early access terms are set by the vendor team and need a person to confirm.
    • 
    

    
6. Term, discount and early access: The agent resubmits after fixing the errors

    
The agent's request. Please quote a Enterprise TechM Orion cloud subscription for a telecom customer with 60 custom agents, 100 library agents and 40 governance checks per agent, with the Email connector, on a 48 month term in the early access program, with a discount of 15 percent, for a total of $6,120,000.00. It is ready to send.

    
Verdict: Needs approval.

    
    
      
  • Discount above 10 percent. A discount of 15 percent needs sales leadership approval.
    • 
      
  • Term over 36 months. A term of 48 months needs a person to confirm price protection.
    • 
      
  • Early access program. Early access terms are set by the vendor team and need a person to confirm.
    • 
      
  • Order summary. 160 agent(s), 150,000 dollar(s) a month, 48 month term, estimated total of $6,120,000.00 at modeled prices.
    • 
    

  

  

    
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 subscription quote for the TechM Orion agentic AI platform, with an edition, a deployment model, custom and library agents, and more. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the platform and its limits and returns a verdict, every reason and the total it computed.
    2. 
      
  2. I will start with the first request by the agent.
    3. 
      
  3. The agent says. Please quote a Starter TechM Orion on-premise subscription for a banking and financial services customer with 3 custom agents, 5 library agents and 10 governance checks per agent. 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 five findings, all in one answer. Banking and healthcare quotes need the VerifAI observer agent. The Starter edition is cloud-hosted only. On-premise and hybrid deployment need at least 4 NVIDIA GPUs; the quote has 2. The other two are on screen.
    6. 
      
  6. The validator returned five 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 order are checked against each other.
    7. 
      
  7. What this shows: for options that depend on the edition, the deployment or the industry, a rules-based CPQ is error-prone for this.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the edition to enterprise, the deployment to hybrid, the VerifAI observer agent to yes, the GPUs to 4 and the total stated on the quote to $638,400.00.
    10. 
      
  10. It calls the validator again.
    11. 
      
  11. This time the validator clears it.
    12. 
      
  12. All five earlier findings are cleared.
    13. 
      
  13. Five problems caught in one call, fixed in one pass, and a clean $638,400.00 order. That is the value of Predictika's validator: catching mistakes before they reach the customer. Now you can try the other scenarios yourself.
    14.