How the Six Defined Archetypes (FDE, SA, PM, LLMOps, Agentic, Transformation) Influence the Career-Ops Job Evaluation Process

Career-Ops uses six role archetypes as a semantic classification layer that determines which CV proof points to prioritize, which narrative templates to apply, and how to calculate the final North-Star alignment score for every job opportunity.

The santifer/career-ops repository employs an archetype-driven evaluation pipeline that begins at Step 0 in modes/oferta.md. By categorizing every job description (JD) into one of six distinct archetypes—AI Forward Deployed Engineer (FDE), Solutions Architect (SA), Technical AI Product Manager (PM), LLMOps Engineer (LLMOps), Agentic Automation (Agentic), and AI Transformation Lead (Transformation)—the system ensures that candidate recommendations are technically relevant and narratively coherent.

Archetype Detection: The Entry Point

The evaluation process starts with archetype detection governed by keyword-based rules defined in modes/_shared.md (lines 78–85). The engine scans the JD for characteristic terms: “client-facing” or “fast delivery” trigger FDE, “architecture” or “integration” trigger SA, “roadmap” or “product manager” trigger PM, “observability” or “evals” trigger LLMOps, “agent” or “orchestration” trigger Agentic, and “transformation” or “adoption” trigger Transformation.

Once detected, the archetype is stored as the foundational classification that drives the remainder of the A–G scoring pipeline documented in modes/oferta.md (lines 5–12).

Three Pillars of Archetype-Driven Evaluation

The detected archetype influences the job evaluation through three interconnected mechanisms: proof-point prioritization, narrative framing, and North-Star alignment scoring.

Proof-Point Prioritisation (Block B – Match with CV)

In Block B, the system maps JD requirements to the candidate’s CV by emphasizing archetype-specific proof points. According to modes/oferta.md (lines 27–33), each archetype dictates a distinct prioritization schema:

• FDE prioritizes delivery speed and client-facing proof points.
• SA prioritizes system design and integrations.
• PM prioritizes product discovery and trade-offs.
• LLMOps prioritizes evals, observability, and pipelines.
• Agentic prioritizes multi-agent orchestration.
• Transformation prioritizes change-management and adoption metrics.

This selective emphasis ensures that the strongest, most relevant evidence surfaces first in the evaluation report.

Narrative and Story Framing (Blocks C, E, and F)

The detected archetype selects the wording templates used for narrative construction. As specified in modes/oferta.md (lines 76–82), the mapping drives language choices across three critical blocks:

1. Block C (Sell Senior Without Lying): Tailors the “sell senior” plan to emphasize archetype-specific competencies—architectural decisions for SA, discovery processes for PM, or delivery speed for FDE.
2. Block E (Customisation Plan): Generates CV and LinkedIn edit lists that highlight achievements aligned with the archetype’s expectations.
3. Block F (STAR+R Stories): Generates interview stories framed around the archetype’s core proof points, ensuring the candidate showcases the most compelling experiences during technical and behavioral interviews.

North-Star Alignment (Score A–F)

The global “North Star alignment” dimension—defined in Section 31–34 of modes/_shared.md—compares the detected archetype against the user-defined target archetypes stored in modes/_profile.md. A close match between the JD’s detected archetype and the candidate’s preferred archetype lifts the overall score, while a mismatch reduces it, ensuring that recommendations align with long-term career trajectory goals.

Implementation in Code

The following Node.js snippet illustrates the core logic used within the Career-Ops pipeline to select proof-point priorities based on the detected archetype:

// src/archetype.js – core logic (illustrative)
import { readFileSync } from 'fs';
import profile from '../modes/_profile.md';

const ARCHETYPE_RULES = {
  FDE:   { proofPoints: ['delivery speed', 'client-facing'] },
  SA:    { proofPoints: ['system design', 'integration'] },
  PM:    { proofPoints: ['product discovery', 'metrics'] },
  LLMOps:{ proofPoints: ['evals', 'observability', 'pipelines'] },
  Agentic:{ proofPoints: ['multi-agent', 'orchestration'] },
  Transformation:{ proofPoints: ['change-management', 'adoption'] },
};

function detectArchetype(jdText) {
  const lower = jdText.toLowerCase();
  if (lower.includes('client-facing') || lower.includes('fast delivery')) return 'FDE';
  if (lower.includes('architecture') || lower.includes('integration')) return 'SA';
  if (lower.includes('roadmap') || lower.includes('product manager')) return 'PM';
  if (lower.includes('observability') || lower.includes('evals')) return 'LLMOps';
  if (lower.includes('agent') || lower.includes('orchestration')) return 'Agentic';
  if (lower.includes('transformation') || lower.includes('adoption')) return 'Transformation';
  return 'UNKNOWN';
}

// Example usage within the evaluation pipeline
const jd = readFileSync('job.txt', 'utf8');
const archetype = detectArchetype(jd);
const proofPoints = ARCHETYPE_RULES[archetype].proofPoints;

// Block B – map JD requirements to CV lines using the selected proof points
const blockB = mapRequirementsToCV(jd, proofPoints, 'cv.md');

In production, this logic is orchestrated through modes/oferta.md (Step 0–4) and modes/_shared.md (Archetype detection table), ensuring that the classification propagates through Blocks B, C, E, and F before final scoring.

Summary

• Archetype classification occurs at Step 0 in modes/oferta.md and acts as the foundational lens for all subsequent evaluation steps.
• Proof-point prioritization in Block B uses archetype-specific criteria to map JD requirements to the strongest CV evidence.
• Narrative framing in Blocks C, E, and F applies archetype-aligned templates for seniority positioning, CV customization, and interview storytelling.
• North-Star alignment compares detected archetypes against user preferences in modes/_profile.md to calculate the final suitability score.
• Keyword-based detection in modes/_shared.md (lines 78–85) enables automatic classification without manual tagging.

Frequently Asked Questions

How does Career-Ops detect which archetype applies to a job description?

The system uses a keyword-based rule set defined in modes/_shared.md (lines 78–85) that scans the JD text for characteristic terms. For example, occurrences of “client-facing” or “fast delivery” classify the role as FDE, while “orchestration” or “multi-agent” classify it as Agentic. This detection runs automatically at Step 0 of the evaluation pipeline before any scoring occurs.

Can I target multiple archetypes in my Career-Ops profile?

Yes. The modes/_profile.md file stores your target archetypes and custom proof-point mappings. The North-Star alignment score (Section 31–34 of modes/_shared.md) compares the JD’s detected archetype against these preferences. A close match increases your overall alignment score, while a mismatch lowers it, helping you filter for roles that fit your career trajectory.

What happens if a job description fits multiple archetypes?

The detection logic in src/archetype.js (and the underlying rules in modes/_shared.md) follows a priority-based sequence. The first matching archetype is selected based on keyword precedence. If no keywords match, the system returns UNKNOWN, and the evaluation pipeline defaults to generic proof-point prioritization until manual classification is provided.

How do the archetypes affect the final PDF or CV output?

The detected archetype influences the templates/ directory (referenced in modes/pdf.md, lines 6–12) to adapt visual emphasis and section ordering. For instance, an LLMOps detection will surface “Evaluations & Observability” sections prominently, while a Transformation detection will prioritize “Change Management” and “Adoption Metrics” in the generated CV layout.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}