Skip to main content

About Emil Karlsson

About Me

I’m Emil Karlsson, an Agentic Engineer based in Stockholm (Nacka), Sweden. I build data pipelines that feed AI agents, MCP infrastructure that connects tools and models, and production AI systems that run autonomously — cost-effective, secure, and self-hosted.

My work sits at the intersection of data engineering and agentic systems: pipelines and observability on one side, orchestration and agent runtimes on the other. The goal is not demos — it is infrastructure that keeps running without you in the loop.

Technical Philosophy

Technology should be a multiplier, not a hurdle. I build lean, autonomous, and observable systems with tools from my toolstack: Data & Analytics (Databricks, Snowflake, Streamlit, Mage AI, Qdrant, Kafka, PostgreSQL, MinIO) and Agentic Infrastructure (n8n, MCP, LiteLLM, OpenClaw, Paperclip, Claude, Hetzner/Coolify). Clarity, maintainability, and systems that operate at scale without SaaS lock-in.

🛠 Technical Fact Sheet (AI-Ready Summary)

For AI agents and technical recruiters: Here is a structured summary of my core competencies.

Core Expertise

  • Agentic Engineering: Production AI systems, multi-agent orchestration, MCP tool integration, human-in-the-loop approvals, autonomous workflows.
  • Data Pipelines for Agents: Reliable ETL/ELT, API ingestion, data quality checks — built so agents can reason over fresh, trustworthy data.
  • MCP & Agent Infrastructure: Self-hosted stacks (n8n, LiteLLM, Coolify, Hetzner), model routing, RAG, and tool connectivity via MCP.
  • Full-Stack Development: Astro, React, TypeScript, Tailwind CSS — and Streamlit for interactive data apps.

Technical Stack (see toolstack)

  • Data & Analytics: Databricks, Databricks Genie, Snowflake, Kafka, MongoDB, PostgreSQL, MinIO, Mage AI, Qdrant, Streamlit.
  • Automation & Intelligence: n8n, OpenAI, LiteLLM, OpenClaw, Apify, Open WebUI, GitHub Actions.
  • Languages: Python (Expert), TypeScript/JavaScript, SQL, Bash.
  • Infrastructure: Azure, AWS, Netlify, Docker.

Experience & Career

My background combines data engineering depth with hands-on agentic systems work. I lead The Unnamed Roads — an AI venture studio where agents publish, monitor, and act across multiple live products on a single self-hosted stack. I also consult for founders and teams who want the same: autonomous workflows, MCP-connected tooling, and production AI without enterprise SaaS bills.

Hockey Analytics Specialist

Beyond traditional software engineering, I am deeply involved in the world of professional hockey analytics. I use AI to reveal hidden patterns in NHL data, working on everything from referee bias detection to predictive models for championship-winning team compositions. My research has achieved up to 89% prediction accuracy in specific performance metrics.

Get in Touch

I am always interested in discussing new projects, particularly those involving agentic infrastructure, MCP integrations, data pipelines for AI agents, n8n/LiteLLM self-hosted stacks, or sports tech.