⌬
Call scorecard vs methodology
EnablementScores a rep's discovery/demo call against your sales methodology rubric (MEDDPICC + iCompaas-specific compliance-deal signals), then returns a per-criterion scorecard with verbatim coachable moments and the single highest-leverage drill for the next 1:1.
The routing recipe
Job typeComplex reasoningPolicyBest qualityConstraintsctx ≥ 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 — 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 — effectively $0 marginal. Speed: ~70 tok/s (estimated).
- Chosen over DeepSeek V4 Pro for cheaper & faster.
| 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 — effectively $0 marginal.
Speed: ~70 tok/s (estimated). vs runner-upChosen over DeepSeek V4 Pro for cheaper & faster. (DeepSeek V4 Pro, score 91.8)
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 ▾
Discovery call, 22 min — AE 'Marcus' x prospect 'Lena, Dir. of Security' at a 90-person health-tech SaaS.
Marcus: Hey Lena, thanks for hopping on — so what made you take the call? Lena: We got a SOC 2 Type II report request from a hospital system we're trying to close, and we have nothing. Marcus: Got it, yeah, a lot of folks come to us that way. So we do SOC 2, ISO, HIPAA — we connect to AWS, Okta, GitHub, auto-collect evidence... [Marcus talks for ~4 min about features] Lena: Okay. We're also looking at Vanta. Marcus: Sure, we're more white-glove than Vanta. Lena: How long does it usually take? Marcus: Depends. I'll send some info over. Lena: Okay sounds good. Marcus: Cool, talk soon!
CALL SCORECARD — Marcus × Lena (health-tech, SOC 2 Type II) — 22 min
1. Agenda/outcome framing — 1/3
"thanks for hopping on — so what made you take the call?" Decent open, but no agenda or stated outcome for the time.
Fix: Frame the 30s up front: "Goal today is to see if we're a fit and, if so, map your path to that hospital report."
2. Compelling event — 3/3
"SOC 2 Type II report request from a hospital system we're trying to close, and we have nothing." Textbook — deal-blocking, dated, external.
3. Quantified pain — 0/3 — NOT OBSERVED
Never asked what the stalled hospital deal is worth, or what 'nothing' costs them in lost ARR/time.
Fix: "What's that hospital contract worth — and what happens to it if the report slips a quarter?"
4. Economic buyer / mobilizer — 0/3 — NOT OBSERVED
Lena is Dir. of Security; no mention of who signs or the deal owner pushing the hospital close.
5. Decision criteria + competitive set — 1/3
"We're also looking at Vanta." Surfaced the competitor but didn't dig: "we're more white-glove" is a claim, not a criterion. No criteria captured.
Fix: "If both could get you the report, what would make you pick one?"
6. Next step with a date — 0/3
"I'll send some info over... talk soon!" No co-created next step, no date. Deal is now drifting.
7. Talk-to-listen — 1/3
~4-min feature monologue right after the pain landed. Rep talked through the most important moment of the call.
OVERALL: 38/100 — strong inbound signal, salvageable, but the rep buried a hot compelling event under a feature dump and exited with no next step.
HIGHEST-LEVERAGE DRILL (next 1:1): Quantify-then-advance. Practice rep — the instant a prospect names a compelling event, ask two questions before pitching: (1) dollarize it, (2) date it. Role-play 3 reps off this exact 'hospital report' opener until Marcus books a dated next step every time.
WHAT GREAT LOOKED LIKE (rewrite of his weakest moment):
Instead of the 4-min feature monologue after "we have nothing," →
"That's the #1 reason teams call us. Before I show you anything — what's that hospital contract worth, and what's the deadline they gave you? ... Okay. So the real job is getting you a credible Type II path that unblocks that deal. Want to put 30 min on the calendar Thursday to map exactly that?"
Test it on your own data