Skip to content

Navigation Menu

Sign in
Sign up
@nkrimmel
nkrimmel
Follow

Nicholas Krimmel nkrimmel

🎯
Focusing

Block or report nkrimmel

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
nkrimmel /README.md

Nicholas Krimmel

Head of Data & AI · Lecturer · Author — Empowering Decisions

I work at the intersection of business, data and technology: 16+ years, most of them in media and publishing, the last eight at C- and director level. I have built cloud data platforms from the ground up on Google Cloud and Azure, taken a double-digit number of AI and generative-AI use cases into production, and set up auditable AI governance under the EU AI Act. My conviction: AI transformation rarely fails on technology, it fails on the organisation. Small, AI-augmented teams with end-to-end ownership are the answer I keep coming back to.

Roles

  • Head of Data & AI at one of Germany's largest trade-media groups — group-wide data and AI function: strategy, platform, analytics, AI assistants and agents for editorial and business processes.
  • Lecturer, FOM University of Applied Sciences, Düsseldorf — modules IT Trends and Innovation and Technology Management and Artificial Intelligence.
  • Author of Klein denken, groß skalieren (2026) — the AI operating system for the German Mittelstand, built on the Microteaming method — and of the dossier series Data & AI im Mittelstand.
  • Advisory — available for advisory, supervisory and expert boards in selected constellations, in particular where technology investments and holdings are decided.

Open source

Practical tools for data and AI leadership, every one runnable, tested and documented. All sample data is synthetic.

AI governance and security

  • eu-ai-act-classifier — classify AI systems under Regulation (EU) 2024/1689: risk class, articles, obligations by role, deadlines; batch mode for AI inventories.
  • shadow-ai-scanner — find unregistered AI usage in codebases and configs (SDKs, endpoints, model IDs, leaked keys, self-hosted servers); SARIF output for CI.
  • prompt-injection-guard — transparent, rule-based guard for prompt injection and data leaks in LLM apps, with canary tokens, PII redaction and an honest benchmark.

Data platform

  • dbt-project-lint — linter for dbt projects that needs neither dbt nor a warehouse: layering, naming, docs, tests, sources, SQL hygiene.
  • publisher-analytics-dbt — a complete local analytics stack for a digital publisher on dbt and DuckDB, including an engagement score and churn features.
  • mcp-duckdb-analyst — MCP server for safe, read-only analytical access to local CSV, Parquet and DuckDB data from Claude and other MCP clients.
  • arxiv-mcp — MCP server for searching and downloading arXiv papers.

AI and machine learning

  • rag-retrieval-eval — offline evaluation of RAG retrieval: chunkers ×ばつ retrievers ×ばつ k with recall@k, MRR, nDCG and failure analysis.
  • subscription-churn-ml — end-to-end churn scoring for subscription businesses: time-based split, calibration, expected-value thresholds, reason codes, model card.
  • llm-tco — total-cost-of-ownership calculator for LLM workloads, hosted APIs vs. self-hosted open-weight models · live.

Knowledge management

  • zweites-gehirn — a German second brain after Andrej Karpathy's LLM-Wiki pattern: immutable raw sources, an agent-maintained knowledge wiki, a schema for Claude Code, plus a CLI for scaffolding, index, lint, search and graph.

Organisation and tooling

  • microteaming-kit — team topology as code: describe your organisation in YAML, lint it against the Microteaming rules, get metrics, dependency maps and impact simulations.
  • yolo-deck, docker-claude-yolo, vibe-yolo — running AI coding agents in isolated containers.

Background

  • M.Sc. Applied Artificial Intelligence, IU International University of Applied Sciences (2026)
  • MBA General Management, Hochschule Fresenius (2024)
  • Certificates: Master Business with AI (MBAI®), Ethics of Artificial Intelligence (LSE), Certified TensorFlow Developer (DeepLearning.AI), Certified Software Engineer (HKUST)

Stack

  • Cloud and data platforms: Google Cloud (BigQuery, Pub/Sub, Cloud Run), Azure, Databricks, Snowflake, DuckDB, dbt, SQL, Python
  • BI and analytics: Looker, Power BI, Tableau, self-service analytics at scale
  • AI: generative AI and LLMs, RAG and embeddings, AI agents, MCP, machine learning for scoring and churn, AI governance under the EU AI Act
  • Engineering and organisation: TypeScript, CI/CD, Kubernetes, Microteaming, agile transformation, OKR

Find me

Popular repositories Loading

  1. nkrimmel nkrimmel Public

    1

  2. arxiv-mcp arxiv-mcp Public

    MCP for the Arxiv Platform

    Python

  3. docker-claude-yolo docker-claude-yolo Public

    Shell

  4. yolo-deck yolo-deck Public

    TypeScript

  5. vibe-yolo vibe-yolo Public

    Shell

  6. dbt-project-lint dbt-project-lint Public

    Fast, warehouse-free linter for dbt projects: layering, naming, docs, tests, sources and SQL hygiene rules with GitHub annotations and SARIF output.

    Python

AltStyle によって変換されたページ (->オリジナル) /