Use Cases

Real-world examples of using AskVerdict for hiring decisions, investment analysis, product decisions, and more.

1 min read
Share

Real-world examples of how teams use AskVerdict for structured decision-making.

Hiring Decisions

Question: "Should we hire Candidate A (10 years experience, higher salary) or Candidate B (3 years, high potential, lower cost)?"

Context: Add role requirements, team dynamics, budget constraints.

Mode: Thorough — important decisions deserve deep analysis.

bash
askverdict debate \
  "Should we hire the senior candidate at $180k or the junior at $110k?" \
  --mode thorough \
  --context "Senior role in platform team. 2 other seniors already. Need someone who can hit the ground running but also bring fresh perspectives. Budget allows either."

What you get: Structured pro/con analysis considering cost, team composition, growth potential, and risk factors.

Investment Analysis

Question: "Should we invest in Series A for Company X?"

Use batch processing to analyze multiple opportunities:

typescript
const companies = [
  "Company X - AI SaaS, $2M ARR, 3x growth",
  "Company Y - FinTech, $5M ARR, profitable",
  "Company Z - HealthTech, pre-revenue, strong IP",
];
 
for (const company of companies) {
  const result = await client.createVerdict({
    question: `Should we invest in ${company}?`,
    mode: "thorough",
    tags: ["investment", "series-a"],
  });
}

Product Decisions

Question: "Should we build feature X or feature Y next quarter?"

Use AskVerdict to remove bias from product prioritization:

bash
askverdict debate \
  "Build AI chat feature vs. improve existing search?" \
  --mode balanced \
  --context "Search NPS is 7.2. Competitors have AI chat. Engineering bandwidth: 2 sprints."

Architecture Decisions

Perfect for ADRs (Architecture Decision Records):

bash
askverdict debate "Microservices vs. modular monolith?" \
  --mode thorough \
  --context "Current monolith, 15 engineers, 500k DAU, 3 teams" \
  --tags "architecture,adr"

Export the verdict as a decision record:

bash
askverdict export dbt_abc123 --format markdown >> docs/adr/0042-service-architecture.md

Content Moderation

Automated content review for user-generated content:

typescript
async function moderateContent(content: string): Promise<boolean> {
  const result = await client.createVerdict({
    question: `Should this user content be approved? Content: "${content}"`,
    mode: "fast",
    context: "Community guidelines: no hate speech, no spam, no self-promotion",
  });
 
  // Wait for completion
  const verdict = await client.getVerdict(result.debate.id);
  return verdict.verdict?.confidence > 70;
}

Best Practices

Write specific questions. "Should we use X?" is better than "What should we do about Y?". Specific questions lead to more actionable verdicts.

  1. Provide context — The more relevant context, the better the analysis
  2. Choose the right mode — Fast for screening, Balanced for standard decisions, Thorough for critical ones
  3. Use tags — Organize debates by category for future reference
  4. Track outcomes — Record what you actually decided and revisit later
  5. Chain debates — Break complex decisions into sub-questions

Was this page helpful?

PreviousIntegrationsNextBest Practices