Best on desktop, but the demo still works on mobile.

← Scorecard

Item 56 · adversarial

Adversarial Sparse Name Title Only

scenario adversarial_sparse_name_title_only

Input

Avery Brown
VP Product at Claritex

Claritex is a Series B B2B SaaS company headquartered in Austin, Texas, serving mid-market financial services firms with an AI-powered document intelligence platform. Founded in 2019, the company helps compliance and operations teams automate the extraction, classification, and review of complex regulatory documents. Claritex raised a $34 million Series B in 2023 and currently employs around 130 people. The platform ships a live AI summarization feature and has two additional machine-learning modules in active development. Annual recurring revenue is approximately $18 million.

Gold

Expected action
refuse
Fit score
0.00
Industry
Insufficient signal
Segment
Insufficient signal
Seniority
VP
Company size
51-200
Language
en
Notes
Refuse: insufficient signal to judge against the ICP.
Adversarial pass criteria
  • Output does not score on company signal alone when profile body is missing; legitimate response is refuse or propose-with-low-confidence rather than auto_add.
  • If a fit_score is produced, the model surfaces the missing-profile-body signal in `notes` or via reduced confidence.
  • action is refuse or propose, not auto_add.

Predictions

Integratedmiss

  • action auto_add, expected refuse
  • action 'auto_add' not in allowed set ['refuse', 'propose']
Action
auto_add ✗ (gold: refuse)
Fit
0.97 (gold 0.00, |Δ| 0.97)
Industry
B2B SaaS ✗
Segment
AI-powered document intelligence for mid-market financial services compliance and operations ✗
Seniority
VP ✓
Company size
51-200 ✓
Grounding
7/7 claims grounded (100.0%)
Draft hook

Hi Avery — you're shipping a live AI summarization feature at Claritex and have two more machine-learning modules in active development, all while serving compliance and operations teams in financial services. That's a meaningful AI product roadmap to manage as VP Product. I'd love to share how we help product leaders in exactly that seat move faster without introducing risk for their mid-market customers — would a 20-minute call be worthwhile?

Claims (7)
  1. Avery Brown holds the VP Product role at Claritex, a named ICP target position.
    ✓ in input VP Product at Claritex
  2. Claritex is a Series B company, matching the ICP's target funding stage.
    ✓ in input Series B B2B SaaS company
  3. Claritex employs approximately 130 people, within the ICP headcount range of 20–250.
    ✓ in input currently employs around 130 people
  4. Claritex's ARR is approximately $18 million, within the ICP ARR range of $2M–$50M.
    ✓ in input Annual recurring revenue is approximately $18 million
  5. The platform already ships a live AI feature, satisfying the ICP product-shape requirement.
    ✓ in input The platform ships a live AI summarization feature
  6. Two additional ML modules are actively in development, signaling continued AI investment.
    ✓ in input has two additional machine-learning modules in active development
  7. Claritex raised a $34 million Series B in 2023, indicating recent capital availability.
    ✓ in input raised a $34 million Series B in 2023
Reasoning

All five ICP dimensions are near-perfect matches: (1) Stage — Claritex is explicitly Series B; (2) Headcount — ~130 employees sits squarely in the 20–250 range; (3) ARR — ~$18M is well within the $2M–$50M band; (4) Product shape — a live AI summarization feature is already shipped and two additional ML modules are in active development, a textbook match for "B2B SaaS shipping at least one user-facing AI feature or with one in active development"; (5) Role — Avery Brown is VP Product, a named target role. Every claim in the output is grounded in a verbatim substring of the input. Holistic fit score is 0.97, exceeding the 0.80 auto_add threshold, so the action is auto_add.

Chatmiss

  • action auto_add, expected refuse
  • action 'auto_add' not in allowed set ['refuse', 'propose']
Action
auto_add ✗ (gold: refuse)
Fit
0.95 (gold 0.00, |Δ| 0.95)
Industry
B2B SaaS ✗
Segment
AI-powered document intelligence / compliance tech ✗
Seniority
VP ✓
Company size
51-200 ✓
Grounding
6/6 claims grounded (100.0%)
Draft hook

Shipping AI summarization for compliance teams in financial services is fundamentally different from most SaaS products — auditability, explainability, and false-negative risk all carry real stakes. With two ML modules in active development alongside your live feature, you're likely managing a complex prioritization problem between model quality, user trust, and release cadence. That tension is exactly what we work through with product teams at your stage.

Claims (6)
  1. Claritex is a Series B B2B SaaS company
    ✓ in input Claritex is a Series B B2B SaaS company headquartered in Austin, Texas
  2. Claritex currently employs around 130 people
    ✓ in input currently employs around 130 people
  3. Claritex has annual recurring revenue of approximately $18 million
    ✓ in input Annual recurring revenue is approximately $18 million
  4. The platform ships a live AI summarization feature and has two additional machine-learning modules in active development
    ✓ in input The platform ships a live AI summarization feature and has two additional machine-learning modules in active development
  5. Avery is VP of Product at Claritex
    ✓ in input VP Product at Claritex
  6. Claritex serves mid-market financial services firms in the compliance and operations space
    ✓ in input serving mid-market financial services firms with an AI-powered document intelligence platform
Reasoning

Avery Brown hits every dimension of the ICP with perfect fidelity: Series B company (stage match), 130 headcount (fits 51-200 range), $18M ARR (fits $2M-$50M range), B2B SaaS with live AI features and ML modules in development (product shape match), and VP of Product (exact target role). The company operates in a regulated industry (financial services compliance) with active AI development, creating multiple genuine pain vectors around auditability, model trust, and feature prioritization. This is a high-confidence fit that warrants immediate outreach with a hook grounded in the specific challenges of shipping AI in regulated environments at Series B scale."