⬣
Outbound cadence architect
OutboundInput a segment/persona plus the channels available; output a benchmark-calibrated multi-touch cadence PLAN — every touch's channel, day-offset, and intent — tuned to ~15.4 attempts over ~29.3 days with a realistic call/email/voicemail/social mix. Designs the structure and timing, not the message copy.
The routing recipe
Job typeComplex reasoningPolicyBalanced
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 — 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 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 — 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 ▾
Segment: cold outbound to Heads of Security / vCISOs at 80-250 employee Series-B SaaS companies whose top customers are starting to demand SOC 2 Type II before renewal. Many are currently on Vanta or have a stalled DIY program. Channels available to our SDRs: call, email, voicemail, LinkedIn. Goal: book a 25-min intro with the AE.
1) CADENCE HEADER
- Segment: Cold — Head of Security / vCISO, Series-B SaaS (80-250 emp), customer-driven SOC 2 Type II pressure, Vanta-installed or stalled DIY.
- Channels used: call, email, voicemail, LinkedIn (all 4 available).
- Total attempts: 15 | Span: Day 1 -> Day 29 (29 days) | Avg spacing: ~1.9 days.
- Channel mix: 5 calls / 5 emails / 3 voicemails / 2 LinkedIn.
2) TOUCH TABLE
Touch | Day | Channel | Intent | Methodology
1 | 1 | Email | Anchor the trigger: top customers now requiring SOC 2 Type II before renewal | CoM current state / SPIN Situation
2 | 1 | Call | First dial — reference the renewal-gating trigger, qualify if this is owned & timed | SPICED Situation
3 | 1 | Voicemail | Paired VM to the unanswered call — name the renewal-revenue-at-risk reason for the call | SPIN Implication
4 | 3 | LinkedIn | Connect; establish credibility on customer-driven SOC 2 deadlines for Series-B SaaS | Sandler up-front (light)
5 | 5 | Call | Surface PAIN: stalled DIY / Vanta evidence gaps vs an auditor-ready Type II window | SPICED Pain
6 | 6 | Email | Quantify IMPACT: cost of a slipped renewal + eng-hours lost to manual evidence collection | SPIN Implication / SPICED Impact
7 | 9 | Call | Introduce the CRITICAL EVENT: map backward from the renewal date to the audit-start date | SPICED Critical Event
8 | 9 | Voicemail | Paired VM — leave the deadline-math hook (audit windows close earlier than buyers expect) | SPICED Critical Event
9 | 12 | Email | Position differentiated capability: continuous evidence + Trust Center vs point-in-time DIY | CoM required capability
10 | 15 | Call | Displacement angle: what Vanta leaves manual; required capabilities a renewal-gated team needs | CoM differentiation
11 | 17 | LinkedIn | Share-style touch: pattern across Series-B peers who hit the same renewal gate (no pitch) | SPIN Need-payoff
12 | 20 | Call | Multi-thread intent: ask for the economic owner of the renewal risk if not this contact | SPICED Decision (stakeholders)
13 | 23 | Voicemail | Paired VM to Touch 12 dial — restate the renewal-clock cost in one line | SPICED Impact
14 | 26 | Email | Need-payoff framing: what an auditor-ready, always-on program unlocks before the next renewal | SPIN Need-payoff / CoM desired outcome
15 | 29 | Email | Graceful break-up: leave the door open, invite a 'circle back at renewal-minus-90' reply | Sandler no-pressure close
3) WEEK-BY-WEEK RHYTHM
- Week 1 (Days 1-7): Highest density — 6 touches. Day-1 triple-tap (email + call + paired VM), LinkedIn connect Day 3, pain-focused call Day 5, impact email Day 6.
- Week 2 (Days 8-14): 3 touches. Pivot to the critical-event / deadline math (call + paired VM Day 9) and position the differentiated capability (email Day 12).
- Week 3 (Days 15-21): 3 touches. Displacement + multi-thread — Vanta-gap call Day 15, peer-pattern LinkedIn Day 17, economic-owner ask Day 20.
- Week 4 (Days 22-29): 3 touches, widening gaps. Paired VM Day 23, need-payoff email Day 26, break-up Day 29.
4) BENCHMARK CHECK
- Attempts: 15 vs ~15.4 target — on benchmark for a cold/blended cadence.
- Span: 29 days vs ~29.3 — on benchmark.
- Spacing: 28 gap-days / ~14 intervals ≈ 1.9-2.0 days avg — on benchmark, with deliberate front-loading (Week 1 = 6 touches) and widening after Week 3, matching real SDR rhythm rather than a flat 1.9-day metronome.
- Mix: 5 call / 5 email / 3 VM / 2 social. Calls + emails are the co-backbone and every VM is paired to an unanswered dial (never standalone), mirroring the aggregate 37:36:15:8 call:email:VM:social blend. Note: this is a cold cadence, so it runs the full ~15/29; an inbound version of this segment would compress to ~10-11 touches over ~18-20 days.
Test it on your own data