SPICED discovery scorer

Use case Β· Sales

← All use cases
⊜

SPICED discovery scorer

Sales

Turn a discovery-call transcript into a SPICED scorecard (Winning by Design) β€” Situation, Pain, Impact, Critical Event, Decision β€” with a per-element 0-3 score, supporting quote, what's missing, an overall deal-readiness score, and the single best next question to fill the biggest gap.

The routing recipe
Job typeComplex reasoningPolicyBest qualityConstraintsctx β‰₯ 128k
routes to
NVIDIA Nemotron 3 Ultra 550B-A55B nvidia
nvidia:nvidia/nemotron-3-ultra-550b-a55b
93.7
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 DeepSeek V4 Pro for cheaper & faster.
Capability the job needsWeightNVIDIA Nemotron 3 Ultra 550B-A55BCatalog rank
reasoning1.00100#1 of 40
math0.6099#3 of 40
knowledge0.5092#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 DeepSeek V4 Pro for cheaper & faster. (DeepSeek V4 Pro, score 91.8)
Example result
Sample input β–Ύ
Discovery call, 26 min β€” AE 'Priya' Γ— prospect 'Tom, VP Engineering' at a 140-person B2B fintech SaaS. Priya: Thanks Tom β€” before we dive in, what's going on that made compliance a priority right now? Tom: We're on AWS, Okta for SSO, GitHub for code. We've never done a formal audit β€” we have a Confluence page of 'policies' and a security questionnaire we fill out by hand every time a prospect asks. Last quarter we got a SOC 2 Type II request from a bank we're trying to land, and we basically froze. We're cobbling evidence together in spreadsheets. Priya: How big is that bank deal, and what happens to it without a report? Tom: It's our biggest pipeline deal β€” around $400k ARR. Their procurement said no report, no contract. They want it before their fiscal year closes, which is March 31. And honestly two other enterprise deals stalled at the security-review stage too. Priya: Who else gets pulled in when you fill those questionnaires today? Tom: Me, a senior eng, and our Head of Security Dana β€” probably 30-plus hours each, every time, and we've done four this quarter. Priya: Got it. If you decided to fix this, who signs off and how do you choose? Tom: Dana owns security and would drive it. Our CFO controls budget. We looked at Vanta's site already. I'd want it live on our AWS and Okta fast. We haven't talked timeline internally yet though.
SPICED SCORECARD β€” Tom @ 140-person fintech SaaS (SOC 2 Type II, bank deal blocked) β€’ Situation (3/3): "We're on AWS, Okta for SSO, GitHub for code... never done a formal audit β€” a Confluence page of 'policies' and a security questionnaire we fill out by hand." | Missing: nothing material β€” stack, posture (no audit), and manual workflow all confirmed. β€’ Pain (3/3): "a SOC 2 Type II request from a bank... we basically froze. We're cobbling evidence together in spreadsheets." Root cause is captured: no connected evidence collection, so every questionnaire is hand-built. | Missing: nothing β€” symptom and root cause both surfaced. β€’ Impact (3/3): "our biggest pipeline deal β€” around $400k ARR... no report, no contract" plus "two other enterprise deals stalled at security-review" and "30-plus hours each... four this quarter" (~120+ eng/security hours/qtr). | Missing: dollar value of the two other stalled deals not yet quantified. β€’ Critical Event (3/3): "They want it before their fiscal year closes, which is March 31." Dated, external, deal-blocking. | Missing: confirm whether a Type II window even fits before Mar 31 (observation period) β€” may force a Type I bridge. β€’ Decision (1/3): "Dana owns security and would drive it. Our CFO controls budget. We looked at Vanta's site already." Stakeholders partly named, Vanta is the competitive set. | Missing: no decision criteria, no internal process/timeline ("haven't talked timeline internally yet"), CFO not engaged, Dana not confirmed as champion. DEAL READINESS: 82/100 β€” Exceptional Pain + Impact + a hard March 31 Critical Event make this a genuinely qualified, urgent opportunity; the only thing standing between this and a forecastable deal is an undefined Decision process and an unengaged economic buyer. BIGGEST GAP: Decision (1/3) β€” the most deal-threatening hole. A $400k deal with a fixed deadline but no agreed criteria, no internal timeline, and a CFO who hasn't been pulled in will stall in approvals even though the buyer wants to move. The Decision element is what converts urgency into a signed contract. BEST NEXT QUESTION: "Since the bank needs the report by March 31 and Dana would drive this, what would the two of you need to see to choose a vendor β€” and can we get 20 minutes with Dana and your CFO this week to back-plan the dates, so we know whether a Type II fits the window or we bridge with a Type I first?"
Test it on your own data
()\nβ€’ situation (x/3): \"\" | missing: <...>\nβ€’ pain (x/3): \"\" | missing: <...>\nβ€’ impact (x/3): \"\" | missing: <...>\nβ€’ critical event (x/3): \"\" | missing: <...>\nβ€’ decision (x/3): \"\" | missing: <...>\ndeal readiness: <0-100> β€” \nbiggest gap: \nbest next question: \"\"","task":"reasoning","optimize":"best_quality","mincontext":128000}'>
Sign in to continue

LLM Switchboard is private β€” sign in with Authlee to access the control room.

Sign in with Authlee
← Back to home