β‘
Methodology question bank builder
SalesInput a prospect/deal context plus ONE chosen sales methodology (CoM, SPICED, SPIN, MEDDIC, BANT, or Sandler); output a discovery question set sequenced through that framework's real stages, with 2-3 rephrasing variants per question and a short example dialogue, tailored to the iCompaas deal.
The routing recipe
Job typeComplex reasoningPolicyBalancedConstraintsctx β₯ 16k
routes to
NVIDIA Nemotron 3 Ultra 550B-A55B nvidia
nvidia:nvidia/nemotron-3-ultra-550b-a55b
90.8
match score
96.5
task fit
100
cost
59.5
speed
Why this model from benchmarks & capabilities
- "Complex reasoning" leans hardest on reasoning; NVIDIA Nemotron 3 Ultra 550B-A55B scores 100/100 there β top-1 of 40 chat models.
- Strength on this class of work shows up on GPQA Diamond, MMLU-Pro, BIG-Bench Hard, where NVIDIA models are competitive.
- Cost: runs on NVIDIA free credits / self-hostable NIM β effectively $0 marginal. Speed: ~70 tok/s (estimated).
- Chosen over NVIDIA Nemotron 3 Super 120B-A12B for higher task-fit (96.5 vs 90.7) & more context.
| Capability the job needs | Weight | NVIDIA Nemotron 3 Ultra 550B-A55B | Catalog rank |
|---|---|---|---|
| reasoning | 1.00 | #1 of 40 | |
| math | 0.60 | #3 of 40 | |
| knowledge | 0.50 | #4 of 40 |
Relevant benchmarksGPQA Diamond MMLU-Pro BIG-Bench Hard LiveBench (reasoning)
EconomicsCost: runs on NVIDIA free credits / self-hostable NIM β effectively $0 marginal.
Speed: ~70 tok/s (estimated). vs runner-upChosen over NVIDIA Nemotron 3 Super 120B-A12B for higher task-fit (96.5 vs 90.7) & more context. (NVIDIA Nemotron 3 Super 120B-A12B, score 89.5)
Speed: ~70 tok/s (estimated). vs runner-upChosen over NVIDIA Nemotron 3 Super 120B-A12B for higher task-fit (96.5 vs 90.7) & more context. (NVIDIA Nemotron 3 Super 120B-A12B, score 89.5)
Example result
Sample input βΎ
Methodology: SPICED
Deal context: Mid-market fintech, ~280 employees, processes card payments. Inbound from a Series B company (Northwind Pay). Champion is Dana Ruiz, newly hired Head of Security (first security hire). They're currently on Vanta but the renewal is in 4 months and Dana is frustrated β says their team still manually chases evidence in spreadsheets and Vanta 'just nags us, doesn't do the work.' A large enterprise customer (their biggest logo) has demanded SOC 2 Type II AND a completed security questionnaire before signing a 7-figure renewal, due end of Q3. Northwind also wants PCI DSS 4.0 by year-end. Dana mentioned the CFO controls budget and is skeptical of 'another tool.'
## Question Bank β SPICED | Northwind Pay
_SPICED fits because there's a clean compelling event (Q3 enterprise renewal gating on SOC 2 Type II + questionnaire) β diagnose Situation through Decision before positioning iCompaas against the Vanta renewal._
### Situation
- Q1: Walk me through how evidence collection actually works today between your team, Vanta, and the spreadsheets you mentioned β who touches what, and how often?
- variant (blunt): What does Vanta automate today vs. what your team still does by hand? | variant (exec): At a high level, where does the compliance program stand against your SOC 2 and PCI commitments? | variant (metrics): How many controls are still manually evidenced each cycle?
- Q2: You're the first dedicated security hire β what did the program look like before you, and what's already on Vanta vs. greenfield?
- variant (blunt): Who owned this before you, and what did you inherit? | variant (exec): How mature would you call the program on a 1-5 scale today? | variant (metrics): How many of the SOC 2 / PCI controls have an owner assigned?
### Pain (problem + root cause)
- Q1: When you say Vanta 'nags but doesn't do the work,' what specifically breaks β is it the integrations not pulling evidence, or the team not trusting what it collects?
- variant (blunt): What's the single most painful part of a Vanta evidence cycle? | variant (exec): Where is the current tooling falling short of what the board expects? | variant (metrics): How many hours/week does your team spend reconciling Vanta against the actual spreadsheets?
- Q2: Is the root cause the tool, the process, or the fact that there's one of you covering SOC 2, PCI, and enterprise questionnaires at once?
- variant (blunt): If you had to name the one thing that has to change, what is it? | variant (exec): Is this a tooling gap or a capacity gap? | variant (metrics): How many open security questionnaires are sitting in your queue right now?
### Impact (quantified cost)
- Q1: If SOC 2 Type II and the questionnaire aren't done by end of Q3, what happens to the enterprise renewal β and what's that logo worth?
- variant (blunt): What's the dollar value of the deal that's gated on this? | variant (exec): What's the revenue exposure if the Q3 date slips? | variant (metrics): 7-figure renewal β is that ARR, TCV, and how much is at risk?
- Q2: Beyond that one renewal, how many deals are stalling in security review, and what's the manual-evidence time costing you in headcount you'd rather spend elsewhere?
- variant (blunt): What is the spreadsheet chase actually costing you? | variant (exec): What's the opportunity cost of your team doing this by hand? | variant (metrics): How many sales cycles per quarter touch a security questionnaire?
### Critical Event (date-bound)
- Q1: The enterprise renewal is end of Q3 and PCI DSS 4.0 is year-end β which is the immovable date, and what has to be true 30 days before it?
- variant (blunt): What's the hard deadline we're working back from? | variant (exec): Which commitment can't slip β the renewal or PCI? | variant (metrics): How many weeks of audit fieldwork do you need before the Q3 date?
- Q2: If we mapped backward from end of Q3, when would a new evidence-automation platform need to be live to make the Type II observation window?
- variant (blunt): How late can you switch off Vanta and still hit Q3? | variant (exec): What's the go-live date that de-risks the renewal? | variant (metrics): What's your Type II observation period β 3, 6, or 12 months?
### Decision (criteria, process, stakeholders)
- Q1: You mentioned the CFO controls budget and is skeptical of 'another tool' β is this a replace-Vanta decision or an add-a-tool decision in their eyes, and who else weighs in?
- variant (blunt): Who has to say yes besides you? | variant (exec): How does a tooling decision of this size get approved here? | variant (metrics): What's the budget envelope already set aside for the Vanta renewal?
- Q2: What would the CFO need to see to treat this as cost-neutral against the Vanta renewal rather than net-new spend?
- variant (blunt): What makes this a no for the CFO? | variant (exec): What's the business case that gets finance on side? | variant (metrics): If we displaced Vanta, what's the renewal number we'd be offsetting?
- Q3: What are your must-have evaluation criteria β integration coverage, questionnaire automation, PCI 4.0 support β and how will you compare us to staying on Vanta?
- variant (blunt): What would make you pick us over renewing Vanta? | variant (exec): What does 'good' look like at the end of the eval? | variant (metrics): How many of your stack's integrations need native evidence pulls?
### Example dialogue
Rep: Walk me through a Vanta evidence cycle today β what's automated vs. by hand?
Buyer (Dana): Honestly, it pulls some cloud config, but my team still chases owners in spreadsheets every quarter. It nags; it doesn't do the work.
Rep: When the enterprise questionnaire landed, how long did it sit before someone could answer it?
Buyer: Three weeks. I'm the only one who can, and I'm buried in SOC 2 prep.
Rep: That renewal β what's gated on getting Type II and the questionnaire done by Q3?
Buyer: It's our biggest logo, seven figures. If we miss it, that's a very bad board meeting.
Rep: So the immovable date is end of Q3, and PCI 4.0 follows at year-end β and the CFO sees this as 'another tool.'
Buyer: Right. He'd need to see it isn't just incremental spend on top of Vanta.
Rep: Then the cleanest path is replacing Vanta so the renewal budget offsets us, with questionnaire automation and PCI 4.0 coverage proven before you commit β that's exactly where iCompaas pulls evidence automatically and auto-drafts the security review. Want to map a Q3-backward timeline?
### Coach's note
- Listen for the dollar figure on the gated renewal and the exact Q3 date β those convert Pain into a quantified, date-bound case the CFO can't wave off.
- Trap question: 'How late can you switch off Vanta and still hit Q3?' β it forces the Critical Event to do the urgency work for you.
- Stop discovering and start positioning only once Decision criteria + the CFO's 'cost-neutral vs. Vanta' frame are explicit; that's the moment differentiated value (auto-evidence, questionnaire automation, PCI 4.0) lands instead of bouncing off as 'another tool.'
Test it on your own data
negative consequences (cost of inaction) -> desired outcomes / required capabilities -> differentiated value (position icompaas only after required capabilities are surfaced). use before/after framing.\n- spiced: situation -> pain (problem + root cause) -> impact (quantified cost) -> critical event (date-bound) -> decision (criteria, process, stakeholders).\n- spin: situation -> problem -> implication (make it hurt) -> need-payoff (value of solving). escalate facts -> value.\n- meddic: metrics -> economic buyer -> decision criteria -> decision process -> identify pain -> champion. generate questions to uncover each (do not score).\n- bant: budget -> authority -> need -> timing. generate questions for each (do not score).\n- sandler: up-front contract (agenda/expectations) -> pain funnel (surface -> business -> personal) -> budget (early) -> decision (process) -> no free consulting. buyer-led.\n\nrules:\n1. honor every stage/letter of the chosen framework, in order, as labeled sections.\n2. per stage, write 2-4 open-ended questions grounded in the specific deal facts (framework deadline, breach, renewal, competitor, stakeholder names, industry). no filler.\n3. for each question give 2-3 rephrasing variants (e.g., a blunter version, a softer/executive version, a metrics-forcing version) so the rep can adapt to the room.\n4. end with a 6-10 line example dialogue that walks one path through the framework with this prospect, showing rep questions and plausible buyer answers, ending where icompaas value naturally enters.\n5. stay methodology-pure: if meddic is chosen, do not slip into spin, etc.\n\noutput format (markdown):\n## question bank β [methodology] | [account]\n_one-line note on how this framework's sequence applies to this deal._\n### [stage 1 name]\n- q1: ...\n - variant (blunt): ... | variant (exec): ... | variant (metrics): ...\n- q2: ... (with variants)\n### [stage 2 name] ... (repeat for every stage in order)\n### example dialogue\nrep: ... / buyer: ... (6-10 turns, ending at value entry)\n### coach's note\n2-3 bullets: what to listen for, the trap question, the moment to stop discovering and start positioning.","task":"reasoning","optimize":"balanced","mincontext":16000}'>