387 lines
17 KiB
JSON
387 lines
17 KiB
JSON
|
{
|
||
|
|
"version": "2026-04-24",
|
||
|
|
"updated": "2026-04-24",
|
||
|
|
"curricula": [
|
||
|
|
{
|
||
|
|
"id": "klassische-ml-metriken",
|
||
|
|
"title": "Klassische ML-Metriken",
|
||
|
|
"short": "ML-Metriken",
|
||
|
|
"icon": "chart",
|
||
|
|
"color": "#0891b2",
|
||
|
|
"description": "Accuracy, Precision, Recall, F1, ROC-AUC, Cohen's Kappa, MCC — welche Metrik wann, und warum",
|
||
|
|
"modules": [
|
||
|
|
{
|
||
|
|
"id": "klassifikation-metriken",
|
||
|
|
"title": "Klassifikations-Metriken",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "acc-prec-rec", "title": "Accuracy vs Precision vs Recall", "objectives": ["Formeln erklären", "Wann welche priorisieren", "Class-Imbalance-Fallstricke erkennen"]},
|
||
|
|
{"id": "f1-mcc-kappa", "title": "F1, MCC, Cohen's Kappa", "objectives": ["Harmonisches Mittel verstehen", "Wann MCC besser als F1", "Inter-Rater-Reliability"]},
|
||
|
|
{"id": "roc-pr-auc", "title": "ROC-AUC vs PR-AUC", "objectives": ["Threshold-unabhängige Bewertung", "Imbalanced Data: warum PR-AUC besser", "Trade-offs visualisieren"]}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "regression-metriken",
|
||
|
|
"title": "Regressions-Metriken",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "rmse-mae", "title": "RMSE vs MAE vs MAPE", "objectives": ["Skalen-Abhängigkeit", "Outlier-Sensitivität", "Relative Fehler"]},
|
||
|
|
{"id": "r2-adjusted", "title": "R² und Adjusted R²", "objectives": ["Erklärte Varianz", "Model-Complexity-Penalty"]}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "nlp-metriken",
|
||
|
|
"title": "NLP & LLM-Metriken",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "bleu-rouge", "title": "BLEU, ROUGE, METEOR", "objectives": ["n-gram-Matching", "Grenzen automatischer Metriken"]},
|
||
|
|
{"id": "perplexity", "title": "Perplexity & Token-Metriken", "objectives": ["Information Theory-Grundlage", "Pro-Token-Verlust"]},
|
||
|
|
{"id": "llm-as-judge", "title": "LLM-as-a-Judge", "objectives": ["Evaluation mit Modell-Richter", "Bias in der Evaluation"]}
|
||
|
|
]
|
||
|
|
}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "business-kpis",
|
||
|
|
"title": "Business-KPIs für KI",
|
||
|
|
"short": "Business-KPIs",
|
||
|
|
"icon": "briefcase",
|
||
|
|
"color": "#0891b2",
|
||
|
|
"description": "ROI, Time-to-Value, Adoption, Kosten-pro-Inference — KI im Unternehmen messbar machen",
|
||
|
|
"modules": [
|
||
|
|
{
|
||
|
|
"id": "roi-tco",
|
||
|
|
"title": "ROI, TCO, Time-to-Value",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "roi-formel", "title": "ROI-Berechnung für KI-Projekte", "objectives": ["FTE-Äquivalente einrechnen", "Indirekte Effekte quantifizieren", "Amortisationsdauer"]},
|
||
|
|
{"id": "tco-hidden", "title": "TCO & versteckte Kosten", "objectives": ["Inferenzkosten vs Trainingskosten", "Operational Overhead", "Vendor-Lock-in"]}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "adoption-satisfaction",
|
||
|
|
"title": "Adoption & Zufriedenheit",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "adoption-rate", "title": "Adoption-Rate & Active-Users", "objectives": ["WAU/MAU-Unterschied", "Benchmarks aus Praxis"]},
|
||
|
|
{"id": "csat-nps", "title": "CSAT, NPS für KI-Features", "objectives": ["Feature-spezifische Messung", "Anti-Patterns"]}
|
||
|
|
]
|
||
|
|
}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "operational",
|
||
|
|
"title": "Operational Metrics",
|
||
|
|
"short": "Operations",
|
||
|
|
"icon": "activity",
|
||
|
|
"color": "#0891b2",
|
||
|
|
"description": "Latency, Throughput, Availability — KI-Systeme im Produktivbetrieb überwachen",
|
||
|
|
"modules": [
|
||
|
|
{
|
||
|
|
"id": "slo-sla",
|
||
|
|
"title": "SLO/SLA-Design",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "latency-percentiles", "title": "P50/P95/P99 Latency", "objectives": ["Warum nicht Average", "Percentile lesen"]},
|
||
|
|
{"id": "throughput-tps", "title": "Throughput & Tokens-per-Second", "objectives": ["LLM-Durchsatz messen", "Batching-Effekte"]},
|
||
|
|
{"id": "availability", "title": "Availability & Error Budget", "objectives": ["SLO-Definition", "Error-Budget-Policy"]}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "monitoring-tooling",
|
||
|
|
"title": "Monitoring-Stack",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "mlflow-arize", "title": "MLflow, Arize, WhyLabs, W&B", "objectives": ["Tool-Landschaft verstehen", "Auswahl-Kriterien"]},
|
||
|
|
{"id": "drift-alerts", "title": "Drift-Alerts & Incident-Playbook", "objectives": ["Schwellenwerte setzen", "False-Positive-Rate"]}
|
||
|
|
]
|
||
|
|
}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "datenqualitaet",
|
||
|
|
"title": "Datenqualität & Drift",
|
||
|
|
"short": "Daten",
|
||
|
|
"icon": "database",
|
||
|
|
"color": "#0891b2",
|
||
|
|
"description": "DAMA-Dimensionen, PSI, KL-Divergence — Datenprobleme vor Modell-Problemen erkennen",
|
||
|
|
"modules": [
|
||
|
|
{
|
||
|
|
"id": "dama-dimensions",
|
||
|
|
"title": "DAMA-Dimensionen",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "completeness-accuracy", "title": "Completeness, Accuracy, Validity", "objectives": ["Jede Dimension mit Beispiel", "Messmethoden"]},
|
||
|
|
{"id": "timeliness-uniqueness", "title": "Timeliness, Uniqueness, Consistency", "objectives": ["Stale-Data-Risiken", "Dedup-Strategien"]}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "drift-detection",
|
||
|
|
"title": "Drift-Detection",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "psi-kl", "title": "PSI, KL- und JS-Divergence", "objectives": ["Formel und Interpretation", "Schwellenwerte"]},
|
||
|
|
{"id": "ks-test", "title": "Kolmogorov-Smirnov-Test", "objectives": ["Nicht-parametrischer Test", "Feature-vs-Label-Drift"]}
|
||
|
|
]
|
||
|
|
}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "bias-fairness",
|
||
|
|
"title": "Bias & Fairness",
|
||
|
|
"short": "Fairness",
|
||
|
|
"icon": "scale",
|
||
|
|
"color": "#0891b2",
|
||
|
|
"description": "Demographic Parity, Equal Opportunity, Disparate Impact — Diskriminierungsfreie Systeme",
|
||
|
|
"modules": [
|
||
|
|
{
|
||
|
|
"id": "fairness-metriken",
|
||
|
|
"title": "Fairness-Metriken",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "demographic-parity", "title": "Demographic Parity", "objectives": ["Formel", "Einschränkungen"]},
|
||
|
|
{"id": "equal-opportunity", "title": "Equal Opportunity & Equalized Odds", "objectives": ["Unterschied zu Parity", "Praxisbeispiel HR-Tool"]},
|
||
|
|
{"id": "disparate-impact", "title": "Disparate Impact & 80%-Regel", "objectives": ["US-EEOC-Standard", "EU-Bezug"]}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "fairness-tooling",
|
||
|
|
"title": "Tooling",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "aif360-fairlearn", "title": "AIF360 & Fairlearn", "objectives": ["Library-Überblick", "Bias-Audits durchführen"]}
|
||
|
|
]
|
||
|
|
}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "explainability",
|
||
|
|
"title": "Explainability (XAI)",
|
||
|
|
"short": "XAI",
|
||
|
|
"icon": "eye",
|
||
|
|
"color": "#0891b2",
|
||
|
|
"description": "SHAP, LIME, Counterfactuals — Modell-Entscheidungen erklärbar machen (EU AI Act Art. 13)",
|
||
|
|
"modules": [
|
||
|
|
{
|
||
|
|
"id": "xai-methoden",
|
||
|
|
"title": "XAI-Methoden",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "shap-lime", "title": "SHAP vs LIME", "objectives": ["Shapley-Values verstehen", "Lokale vs globale Erklärung"]},
|
||
|
|
{"id": "counterfactuals", "title": "Counterfactual & Anchor Explanations", "objectives": ["Minimal-Änderungs-Prinzip", "Praxisnutzen"]}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "xai-qualitaet",
|
||
|
|
"title": "Qualität von Erklärungen",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "fidelity-stability", "title": "Fidelity, Stability, Comprehensibility", "objectives": ["Messgrößen für Erklärungsqualität", "Trade-offs"]}
|
||
|
|
]
|
||
|
|
}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "robustheit-security",
|
||
|
|
"title": "Robustheit & Security",
|
||
|
|
"short": "Robustheit",
|
||
|
|
"icon": "shield",
|
||
|
|
"color": "#0891b2",
|
||
|
|
"description": "Adversarial Robustness, Prompt-Injection, Data-Poisoning — Angriffsvektoren verstehen",
|
||
|
|
"modules": [
|
||
|
|
{
|
||
|
|
"id": "adversarial",
|
||
|
|
"title": "Adversarial Robustness",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "pgd-fgsm", "title": "PGD & FGSM-Attacks", "objectives": ["Perturbations verstehen", "L-infinity-Budget"]},
|
||
|
|
{"id": "certified-robustness", "title": "Certified Robustness", "objectives": ["Formale Garantien vs empirische Tests"]}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "llm-security",
|
||
|
|
"title": "LLM-Security",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "prompt-injection", "title": "Prompt-Injection & Jailbreaks", "objectives": ["OWASP LLM Top 10", "Mitigationen"]},
|
||
|
|
{"id": "data-poisoning", "title": "Data-Poisoning-Erkennung", "objectives": ["Training-Set-Forensik", "Monitoring-Metriken"]}
|
||
|
|
]
|
||
|
|
}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "governance-reifegrad",
|
||
|
|
"title": "Governance & Reifegrad",
|
||
|
|
"short": "Reifegrad",
|
||
|
|
"icon": "gauge",
|
||
|
|
"color": "#0891b2",
|
||
|
|
"description": "Gartner AI Maturity, MIT CISR, Microsoft RAI MM — wo stehst du, wo willst du hin",
|
||
|
|
"modules": [
|
||
|
|
{
|
||
|
|
"id": "reifegradmodelle",
|
||
|
|
"title": "AI-Reifegradmodelle im Vergleich",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "gartner-maturity", "title": "Gartner AI Maturity Model", "objectives": ["5 Stufen", "Self-Assessment"]},
|
||
|
|
{"id": "mit-cisr", "title": "MIT CISR & Microsoft RAI MM", "objectives": ["Unterschiede", "DACH-Anwendung"]}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "self-assessment",
|
||
|
|
"title": "Self-Assessment-Dimensionen",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "strategy-data", "title": "Strategy, Data, Technology", "objectives": ["Fragebogen-Struktur", "Reifegrad-Score berechnen"]},
|
||
|
|
{"id": "people-processes", "title": "People, Processes", "objectives": ["Kompetenz-Matrix", "Prozess-Reife"]}
|
||
|
|
]
|
||
|
|
}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "eu-ai-act",
|
||
|
|
"title": "EU AI Act Compliance",
|
||
|
|
"short": "EU AI Act",
|
||
|
|
"icon": "flag",
|
||
|
|
"color": "#0891b2",
|
||
|
|
"description": "Risikoklassen, CE-Kennzeichnung, Artikel-Pflichten — Was muss ich ab wann erfüllen",
|
||
|
|
"modules": [
|
||
|
|
{
|
||
|
|
"id": "risikoklassen",
|
||
|
|
"title": "Risikoklassen",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "verboten-high-risk", "title": "Verboten vs High-Risk", "objectives": ["Art. 5 vs Annex III", "Grenzfälle erkennen"]},
|
||
|
|
{"id": "limited-minimal", "title": "Limited & Minimal Risk", "objectives": ["Transparenz-Pflichten", "Chatbots"]}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "high-risk-pflichten",
|
||
|
|
"title": "High-Risk-Pflichten",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "risk-management", "title": "Art. 9 Risikomanagement-System", "objectives": ["Dokumentations-Anforderungen"]},
|
||
|
|
{"id": "data-governance", "title": "Art. 10 Data Governance", "objectives": ["Trainings-, Validierungs- und Testdaten"]},
|
||
|
|
{"id": "transparency-art13", "title": "Art. 13 Transparenz", "objectives": ["User-Information", "Logging"]},
|
||
|
|
{"id": "human-oversight", "title": "Art. 14 Menschliche Aufsicht", "objectives": ["Override-Mechanismen"]}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "timeline-ce",
|
||
|
|
"title": "Timeline & CE-Kennzeichnung",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "phasen-einfuehrung", "title": "Inkrafttreten-Phasen", "objectives": ["2025/2026/2027-Meilensteine"]},
|
||
|
|
{"id": "ce-marking", "title": "CE-Kennzeichnung für High-Risk", "objectives": ["Konformitäts-Assessment", "Benannte Stellen"]}
|
||
|
|
]
|
||
|
|
}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "nist-ai-rmf",
|
||
|
|
"title": "NIST AI RMF",
|
||
|
|
"short": "NIST",
|
||
|
|
"icon": "book",
|
||
|
|
"color": "#0891b2",
|
||
|
|
"description": "Govern, Map, Measure, Manage — US-Framework mit Crosswalk zu EU AI Act und ISO 42001",
|
||
|
|
"modules": [
|
||
|
|
{
|
||
|
|
"id": "vier-funktionen",
|
||
|
|
"title": "Die 4 Kernfunktionen",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "govern", "title": "Govern", "objectives": ["Organisations-Kultur", "Policies"]},
|
||
|
|
{"id": "map", "title": "Map", "objectives": ["Kontext & Risiken erfassen"]},
|
||
|
|
{"id": "measure", "title": "Measure", "objectives": ["KPIs und Tests"]},
|
||
|
|
{"id": "manage", "title": "Manage", "objectives": ["Priorisieren, Response, Recovery"]}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "crosswalk",
|
||
|
|
"title": "Crosswalk zu anderen Frameworks",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "nist-vs-eu-ai-act", "title": "NIST vs EU AI Act", "objectives": ["Überlappungen nutzen"]},
|
||
|
|
{"id": "nist-vs-iso", "title": "NIST vs ISO 42001", "objectives": ["Mapping-Tabellen"]}
|
||
|
|
]
|
||
|
|
}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "iso-42001",
|
||
|
|
"title": "ISO/IEC 42001 & 23894",
|
||
|
|
"short": "ISO",
|
||
|
|
"icon": "award",
|
||
|
|
"color": "#0891b2",
|
||
|
|
"description": "AI Management System, AI Risk Management — zertifizierbare Standards seit 2023",
|
||
|
|
"modules": [
|
||
|
|
{
|
||
|
|
"id": "iso-42001-aims",
|
||
|
|
"title": "ISO 42001 — AIMS",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "aims-struktur", "title": "Was ist ein AIMS", "objectives": ["Management-System-Struktur", "PDCA für KI"]},
|
||
|
|
{"id": "zertifizierungspfad", "title": "Zertifizierungspfad", "objectives": ["Akkreditierte Stellen DACH", "Aufwand schätzen"]}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "iso-23894",
|
||
|
|
"title": "ISO 23894 — AI Risk Management",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "risk-assessment", "title": "Risk Assessment Prozess", "objectives": ["Risiko identifizieren/analysieren/bewerten"]},
|
||
|
|
{"id": "komplement-27001", "title": "Komplementarität zu ISO 27001", "objectives": ["Synergie mit ISMS"]}
|
||
|
|
]
|
||
|
|
}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "scorecards",
|
||
|
|
"title": "Scorecards & Dashboards",
|
||
|
|
"short": "Scorecards",
|
||
|
|
"icon": "clipboard",
|
||
|
|
"color": "#0891b2",
|
||
|
|
"description": "KPI-Hierarchie Exec/Operational/Technical — vom Dashboard zur Entscheidung",
|
||
|
|
"modules": [
|
||
|
|
{
|
||
|
|
"id": "scorecard-design",
|
||
|
|
"title": "Scorecard-Design",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "kpi-hierarchie", "title": "Exec-, Operational-, Technical-Layer", "objectives": ["Welche KPI auf welcher Ebene", "Aggregations-Logik"]},
|
||
|
|
{"id": "vanity-metrics", "title": "Vanity-Metrics erkennen", "objectives": ["Anti-Patterns", "Actionability-Test"]}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "dashboarding-tools",
|
||
|
|
"title": "Dashboarding-Tools",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "tableau-looker-superset", "title": "Tableau, Looker, Superset, Grafana", "objectives": ["Stärken/Schwächen", "Kosten"]}
|
||
|
|
]
|
||
|
|
}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "okr-ki",
|
||
|
|
"title": "OKRs für KI-Teams",
|
||
|
|
"short": "OKR",
|
||
|
|
"icon": "target",
|
||
|
|
"color": "#0891b2",
|
||
|
|
"description": "Objectives & Key Results — Balance zwischen Tech-KPIs und Business-Impact",
|
||
|
|
"modules": [
|
||
|
|
{
|
||
|
|
"id": "okr-grundlagen",
|
||
|
|
"title": "OKR-Grundlagen für AI",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "obj-vs-kr", "title": "Objectives vs Key Results", "objectives": ["Qualitativ vs quantitativ"]},
|
||
|
|
{"id": "balance-tech-biz", "title": "Balance Tech vs Business", "objectives": ["Anti-Pattern: Over-Tech-Indexed"]}
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"id": "beispiele-dach",
|
||
|
|
"title": "Beispiele aus DACH-Mittelstand",
|
||
|
|
"subtopics": [
|
||
|
|
{"id": "mittelstand-cases", "title": "5 Praxisbeispiele", "objectives": ["Formulierungs-Muster"]}
|
||
|
|
]
|
||
|
|
}
|
||
|
|
]
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"badges": [
|
||
|
|
{"id": "erste_metrik", "title": "Erste Metrik-Analyse", "icon": "target", "description": "1. Quiz zu klassischen ML-Metriken bestanden"},
|
||
|
|
{"id": "metrik_master", "title": "Metrik-Master:in", "icon": "trophy", "description": "10 Klassifikations-Metriken-Fragen korrekt"},
|
||
|
|
{"id": "bias_hunter", "title": "Bias-Jäger:in", "icon": "scale", "description": "Bias & Fairness-Modul abgeschlossen"},
|
||
|
|
{"id": "ai_act_navigator", "title": "EU-AI-Act-Navigator:in", "icon": "flag", "description": "EU-AI-Act-Modul abgeschlossen"},
|
||
|
|
{"id": "nist_practitioner", "title": "NIST-Praktiker:in", "icon": "book", "description": "NIST AI RMF-Modul abgeschlossen"},
|
||
|
|
{"id": "iso_expert", "title": "ISO-42001-Expert:in", "icon": "award", "description": "ISO 42001/23894-Modul abgeschlossen"},
|
||
|
|
{"id": "scorecard_architect", "title": "Scorecard-Architekt:in", "icon": "clipboard", "description": "Scorecards-Flashcards bestanden"},
|
||
|
|
{"id": "governance_lead", "title": "Governance-Lead", "icon": "crown", "description": "Reifegrad-Modul abgeschlossen"},
|
||
|
|
{"id": "kpi_master", "title": "KI-Kennzahlen-Master", "icon": "star", "description": "Alle 13 Curricula abgeschlossen"},
|
||
|
|
{"id": "night_owl", "title": "Nachteule", "icon": "moon", "description": "Nach 22 Uhr gelernt"},
|
||
|
|
{"id": "early_bird", "title": "Frühaufsteher:in", "icon": "sun", "description": "Vor 7 Uhr gelernt"}
|
||
|
|
],
|
||
|
|
"levels": [
|
||
|
|
{"min": 0, "title": "Einsteiger:in"},
|
||
|
|
{"min": 50, "title": "Analyst:in"},
|
||
|
|
{"min": 200, "title": "Data-Scientist:in"},
|
||
|
|
{"min": 500, "title": "ML-Engineer:in"},
|
||
|
|
{"min": 1250, "title": "AI-Program-Lead"},
|
||
|
|
{"min": 2500, "title": "Head of AI"},
|
||
|
|
{"min": 5000, "title": "Chief Data/AI-Officer"}
|
||
|
|
]
|
||
|
|
}
|