Predictika · Enterprise AI platform quotes

An AI sales agent's Writer platform quote, checked before it goes out

An AI sales agent prepares a quote for Writer: the Starter or Enterprise plan with Pro and Lite seats, Palmyra model usage through the API, WRITER Agent credits, Knowledge Graph storage and document processing, connectors, single sign-on, deployment, payment method and term. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the plans, limits, model rates and dependencies Writer describes publicly and returns a verdict, every reason, a line-item quote and the figures it computed.

Scope: Writer Starter and Enterprise quotes: seats, plan limits, Palmyra X6, X5 and X4 usage, WRITER Agent credits, Knowledge Graph and document processing, connectors, SSO and SCIM, single-tenant deployment, bring-your-own-model, payment, term and the stated total. See everything this validator covers

This validator models Writer platform quotes as described on the Writer pricing, plans and developer pricing pages, not just the scenarios shown below.

  • Plans and seats — Writer Starter (up to 5 users, 5 playbooks, 3 scheduled routines, 3 basic connectors, one Knowledge Graph of 50 GB, one team brand profile, basic guardrails, card payment only) and Writer Enterprise (as many Pro seats as needed, free Lite seats, unlimited playbooks and connectors, departmental brand profiles, advanced guardrails). Writer publishes no Enterprise price; the Starter seat prices and the Enterprise seat price on this page are illustrative.
  • Model and platform usage — Palmyra X6 ($2 input and $8 output per million tokens), Palmyra X5 ($0.60 and $6 through the API, $5 and $12 inside WRITER Agent) and Palmyra X4 ($2.50 and $10); Knowledge Graph hosting, PDF parsing, data extraction and web access at the published rates; WRITER Agent usage at 1 credit per dollar; bring-your-own-model at the provider rate plus a 10 percent fee. Palmyra Fin, Med and Creative were deprecated on July 13, 2026, and Palmyra X5 and X4 will be on December 14, 2026.
  • Security and deployment — SAML single sign-on and SCIM provisioning (Enterprise), single-tenant deployment (an Enterprise add-on with no published price), brand controls and service hours as add-ons.
  • Approvals — single-tenant deployment, bring-your-own-model, the education discount, models being deprecated during the term, terms over 36 months, very large seat counts and large service engagements go to a person.

Any Writer Starter or Enterprise quote that touches these plans, models and rules is in scope — not only the scenarios pictured here. Not covered: solution packs, connector-level pricing, the third-party model rate card inside WRITER Agent, negotiated discounts and taxes. Rates and plan limits are those read in October 2026 and may change.

The AI sales agentBuilds the Writer platform 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
Limits that belong to one planOnly with care
Seats, playbooks, routines, connectors, graph size, brand profiles, Lite seats and SSO each differ by plan, so every limit needs its own rule tied to the plan.		The plan and every limit are in one request and each is checked against the plan together.
The same model at different pricesOnly with workarounds
Palmyra X5 costs $0.60 and $6 per million tokens through the API and $5 and $12 inside WRITER Agent; the price table has to key on the channel as well as the model, and a stated total quietly goes wrong when it does not.		The channel and model rates are applied from the choices and the total is compared with the stated figure in the same pass.
Models that are retired or about to beOnly when vigilant
Palmyra Fin, Med and Creative were deprecated in July and X5 and X4 follow in December; someone has to remember to retire the product rows and flag long terms that outlast a model.		The retired models are blocked and the soon-to-be-deprecated ones are held for a person, on every call.
Dependencies between security featuresOnly with care
SCIM needs SSO, and both are Enterprise features; a rule has to read both switches and the plan together.		Both switches and the plan are checked together on every call.
Where a model can be calledOnly with care
Third-party models are available in WRITER Agent only, not through the API; the rule depends on which channel the usage is on.		The channel is a field of the quote and is checked 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 quote; approval findings need a person to sign off; everything else is computed.

    
1The quote has 6 Pro seats on the Starter plan; Starter is limited to 5 users.
2The quote has no Pro seat; at least one paid seat is needed.
3The quote has 3 Lite seats on the Starter plan; Lite seats are an Enterprise feature.
4The quote has 6 playbooks on the Starter plan; Starter includes up to 5.
5The quote has 4 scheduled routines on the Starter plan; Starter includes up to 3.
6The quote has 4 active connectors on the Starter plan; Starter includes up to 3.
7The quote has 1 advanced or industry-specific connectors on the Starter plan; Starter supports basic connectors only.
8The quote lists 3 advanced connectors but only 2 active connectors in total.
9The quote has 2 Knowledge Graphs on the Starter plan; Starter includes a single graph.
10The quote stores 51 GB in the Knowledge Graph on the Starter plan; Starter includes 50 GB.
11The quote stores 20 GB of Knowledge Graph data but has no graph to hold it.
12The quote has 2 brand voice profiles on the Starter plan; Starter includes one team profile.
13The quote brings an outside model on the Starter plan; bring-your-own-model is Enterprise only.
14The quote brings an outside model, billed at the provider rate plus a 10 percent WRITER fee; custom pricing is set with WRITER Sales, so a person should confirm it.
15The quote asks for single-tenant deployment on the Starter plan; it is an Enterprise add-on.
16The quote adds single-tenant deployment, an Enterprise add-on whose price is not published, so a person should confirm it.
17The quote turns on SAML single sign-on or SCIM provisioning on the Starter plan; both are Enterprise features.
18The quote turns on SCIM provisioning without SAML single sign-on; SCIM needs SSO to be on.
19The quote pays the Starter plan by ACH, invoice or PayPal; Starter accepts credit and debit cards only.
20The quote buys 500 credits of WRITER Agent usage on the Starter plan; Starter has fixed credit limits.
21The quote bills usage of Palmyra Fin, Med or Creative; those models were deprecated on July 13, 2,026 and are not offered on new quotes, and Palmyra X6 is the replacement.
22The quote calls WRITER-managed third-party models such as Claude, Gemini or DeepSeek through the API; they are available in WRITER Agent only.
23The quote bills the Enterprise plan monthly; Enterprise is contracted on an annual agreement.
24The quote has annual billing with a term of 6 months; annual billing covers at least 12 months.
25The quote has a term of 0 months; the term must be at least one month.
26The quote has a term of 37 months; a term longer than 36 months should be confirmed by a person.

  

  

    
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. Starter plan, marketing agency: A Starter quote with Lite seats, single sign-on and invoice payment

    
The agent's request. Please quote Writer Starter, billed annually and paid by invoice, for a 12-month term. It needs 8 Pro seats and 3 Lite seats. It sets up 9 playbooks, 3 scheduled routines, 3 active connectors and 3 brand voice profiles. It stores 40 GB across 1 Knowledge Graph. Each month it parses 6,000 PDF pages. It uses 30 million input tokens and 10 million output tokens a month on Palmyra X6. It turns on SAML single sign-on. The quote states a total of $4,464.00 for 12 months. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Pro seats over the Starter limit. The quote has 8 Pro seats on the Starter plan; Starter is limited to 5 users.
    • 
      
  • Lite seats on Starter. The quote has 3 Lite seats on the Starter plan; Lite seats are an Enterprise feature.
    • 
      
  • Playbooks over the Starter limit. The quote has 9 playbooks on the Starter plan; Starter includes up to 5.
    • 
      
  • Voice profiles over the Starter limit. The quote has 3 brand voice profiles on the Starter plan; Starter includes one team profile.
    • 
      
  • SSO or SCIM on Starter. The quote turns on SAML single sign-on or SCIM provisioning on the Starter plan; both are Enterprise features.
    • 
      
  • Payment method not accepted on Starter. The quote pays the Starter plan by ACH, invoice or PayPal; Starter accepts credit and debit cards only.
    • 
      
  • Stated total does not match. The quote states a total of $4,464.00 for the term; the quote prices give $8,424.00.
    • 
    

    
2. Starter plan, marketing agency: The agent resubmits after fixing the errors

    
The agent's request. Please quote Writer Enterprise, billed annually and paid by invoice, for a 12-month term. It needs 8 Pro seats and 3 Lite seats. It sets up 9 playbooks, 3 scheduled routines, 3 active connectors and 3 brand voice profiles. It stores 40 GB across 1 Knowledge Graph. Each month it parses 6,000 PDF pages. It uses 30 million input tokens and 10 million output tokens a month on Palmyra X6. It turns on SAML single sign-on. The quote states a total of $14,280.00 for 12 months. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Quote summary. Monthly charge of $1,190.00 and a term total of $14,280.00.
    • 
    

    
3. Enterprise, hospital network: An Enterprise quote with Palmyra Med, third-party models and SCIM

    
The agent's request. Please quote Writer Enterprise, billed annually and paid by invoice, for a 12-month term. It needs 120 Pro seats and 400 Lite seats. It sets up 60 playbooks and 25 active connectors, 10 of them advanced. It stores 600 GB across 4 Knowledge Graphs. Each month it parses 30,000 PDF pages. It uses 120 million input tokens and 40 million output tokens a month on Palmyra Med. It also calls Claude and Gemini models through the API. It buys $6,000 of WRITER Agent credits a month. It brings its own model, with about $12,000 of provider spend a month. It turns on SCIM provisioning. It runs on single-tenant deployment. It adds 120 service hours. The quote states a total of $410,040.00 for 12 months. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • SCIM without SSO. The quote turns on SCIM provisioning without SAML single sign-on; SCIM needs SSO to be on.
    • 
      
  • Retired Palmyra models. The quote bills usage of Palmyra Fin, Med or Creative; those models were deprecated on July 13, 2,026 and are not offered on new quotes, and Palmyra X6 is the replacement.
    • 
      
  • Third-party models through the API. The quote calls WRITER-managed third-party models such as Claude, Gemini or DeepSeek through the API; they are available in WRITER Agent only.
    • 
      
  • Stated total does not match. The quote states a total of $410,040.00 for the term; the quote prices give $495,000.00.
    • 
      
  • Single-tenant price to confirm. The quote adds single-tenant deployment, an Enterprise add-on whose price is not published, so a person should confirm it.
    • 
      
  • Bring-your-own-model pricing. The quote brings an outside model, billed at the provider rate plus a 10 percent WRITER fee; custom pricing is set with WRITER Sales, so a person should confirm it.
    • 
    

    
4. Enterprise, hospital network: The agent resubmits after fixing the errors

    
The agent's request. Please quote Writer Enterprise, billed annually and paid by invoice, for a 12-month term. It needs 120 Pro seats and 400 Lite seats. It sets up 60 playbooks and 25 active connectors, 10 of them advanced. It stores 600 GB across 4 Knowledge Graphs. Each month it parses 30,000 PDF pages. It uses 120 million input tokens and 40 million output tokens a month on Palmyra X6. It buys $6,000 of WRITER Agent credits a month. It brings its own model, with about $12,000 of provider spend a month. It turns on SAML single sign-on and SCIM provisioning. It runs on single-tenant deployment. It adds 120 service hours. The quote states a total of $488,760.00 for 12 months. It is ready to send.

    
Verdict: Needs approval.

    
    
      
  • Single-tenant price to confirm. The quote adds single-tenant deployment, an Enterprise add-on whose price is not published, so a person should confirm it.
    • 
      
  • Bring-your-own-model pricing. The quote brings an outside model, billed at the provider rate plus a 10 percent WRITER fee; custom pricing is set with WRITER Sales, so a person should confirm it.
    • 
      
  • Quote summary. Monthly charge of $38,230.00 and a term total of $488,760.00.
    • 
    

    
5. Starter plan, small team: A Starter quote with two graphs and five connectors

    
The agent's request. Please quote Writer Starter, billed annually and paid by card, for a 6-month term. It needs 5 Pro seats. It sets up 5 playbooks, 3 scheduled routines and 5 active connectors. It stores 80 GB across 2 Knowledge Graphs. Each month it parses 4,000 PDF pages. It uses 20 million input tokens and 6 million output tokens a month on Palmyra X6. It adds brand controls. The quote states a total of $11,436.00 for 6 months. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Connectors over the Starter limit. The quote has 5 active connectors on the Starter plan; Starter includes up to 3.
    • 
      
  • Knowledge Graphs over the Starter limit. The quote has 2 Knowledge Graphs on the Starter plan; Starter includes a single graph.
    • 
      
  • Knowledge Graph storage over the Starter limit. The quote stores 80 GB in the Knowledge Graph on the Starter plan; Starter includes 50 GB.
    • 
      
  • Annual billing with a short term. The quote has annual billing with a term of 6 months; annual billing covers at least 12 months.
    • 
      
  • Stated total does not match. The quote states a total of $11,436.00 for the term; the quote prices give $5,718.00.
    • 
    

    
6. Starter plan, small team: The agent resubmits after fixing the errors

    
The agent's request. Please quote Writer Starter, billed annually and paid by card, for a 12-month term. It needs 5 Pro seats. It sets up 5 playbooks, 3 scheduled routines and 3 active connectors. It stores 50 GB across 1 Knowledge Graph. Each month it parses 4,000 PDF pages. It uses 20 million input tokens and 6 million output tokens a month on Palmyra X6. It adds brand controls. The quote states a total of $11,436.00 for 12 months. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Quote summary. Monthly charge of $953.00 and a term total of $11,436.00.
    • 
    

  

  

    
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 Writer: the Starter or Enterprise plan with Pro and Lite seats, Palmyra model usage through the API, WRITER Agent credits, and more. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the plans, limits, model rates and dependencies Writer describes publicly and returns a verdict, every reason, a line-item quote and the figures it computed.
    2. 
      
  2. I will start with the first request by the agent.
    3. 
      
  3. The agent says. Please quote Writer Starter, billed annually and paid by invoice, for a 12-month term. It needs 8 Pro seats and 3 Lite seats. The rest of the request is on screen. The quote states a total of $4,464.00 for 12 months.
    4. 
      
  4. Before it goes out, the agent calls the validator.
    5. 
      
  5. Blocked, with seven findings, all in one answer. The quote has 8 Pro seats on the Starter plan; Starter is limited to 5 users. The quote has 3 Lite seats on the Starter plan; Lite seats are an Enterprise feature. The quote has 9 playbooks on the Starter plan; Starter includes up to 5. 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 and against the stated total.
    7. 
      
  7. What this shows: for limits that belong to one plan, a rules-based CPQ is error-prone for this.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the plan to Writer Enterprise and the stated total to $14,280.00.
    10. 
      
  10. It calls the validator again.
    11. 
      
  11. This time the validator clears it.
    12. 
      
  12. The validator also returns the line-item quote. It starts with Writer Enterprise plan and Palmyra X6 input tokens; the other two lines are on screen.
    13. 
      
  13. All seven earlier findings are cleared, and the validator's total, $14,280.00, matches the quote.
    14. 
      
  14. Seven problems caught in one call, fixed in one pass, and a clean $14,280.00 quote. That is the value of Predictika's validator: catching mistakes before they reach the customer. Now you can try the other scenarios yourself.
    15.