Lead Qualification Workflow

A 3-step Temporal workflow that enriches lead data, scores leads using AI (0-100), and routes them to the appropriate GTM motion.

Workflow Definition

interface LeadQualificationInput {
  leadId: string;
  company: string;
  contact: string;
  source: string;
}
 
interface LeadQualificationOutput {
  score: number;        // 0-100
  route: string;        // 'ae-direct' | 'bdr-nurture' | 'marketing-drip'
  enrichedData: object;
  reasoning: string;
}

Steps

Enrich Lead Data Activity: enrichLeadData

Gathers additional data about the lead:

• Company firmographics (size, industry, funding, tech stack)
• Contact role and seniority
• Recent activity signals (website visits, content downloads)
• Existing relationship context from memory

AI Scoring Activity: scoreLeadWithAI Model: Claude CLI

AI-powered scoring using the enriched data. The model evaluates against qualification criteria (BANT + MEDDIC hybrid) and returns a score from 0-100 with reasoning.

const result = await scoreLeadWithAI({
  enrichedData,
  scoringCriteria: await readMemory('org', 'qualification-criteria'),
  historicalPatterns: await readMemory('team', 'scoring-baselines')
});
// Returns: { score: 82, reasoning: "Strong ICP fit, budget confirmed, ..." }

Route Lead Activity: routeLead

Routes the lead based on score:

Queries

Scoring Criteria

The AI scoring model evaluates across these dimensions:

Memory Integration

• Scoring criteria stored in org scope (shared across all workers)
• Historical scoring baselines stored in team scope
• Individual lead context stored in worker scope for the assigned BDR