Data Integration · AI Agent Integration · Applied ML
📍 Cologne, Germany 🇩🇪
Growth never stops. I stay ahead of the curve by adapting to industry trends and emerging technologies.
I connect systems, data and AI so they work together in production:
- 🔄 Data integration: connectors, ETL/feed pipelines and format transformation (CSV ⇄ XML, marketplaces, catalogues, APIs)
- 🤖 AI agent integration: giving agents safe, well-defined access to real systems through protocols, tool interfaces and authorization controls
- 🧠 Applied ML: turning models (e.g. transformers) into practical solutions such as classification and categorization on real business data
| Project | What it does | Stack |
|---|---|---|
| ucp-connector-for-woocommerce | Implements the Universal Commerce Protocol (UCP) and ACP so AI agents can transact with WooCommerce, with secure checkout session APIs and agent authorization controls | PHP |
| enterprise-agentic-playbook | Open-source playbook for adopting agentic AI in the enterprise (2026) | Docs |
| codex-cstack | Turns Codex from one generic assistant into explicit delivery modes (planning, engineering design, review, release, QA, retros), inspired by gstack | Shell |
| openclaw-oracle-starter | One-script installer for a self-hosted OpenClaw server on an Oracle Cloud Ubuntu VM: Gateway, Open WebUI, HTTPS via Nginx and Let's Encrypt, Telegram access, web search and browser fallback | Shell |
| Project | What it does | Stack |
|---|---|---|
| Mirakl-Woocommerce-Connector | Syncs products and orders between WooCommerce and the Mirakl marketplace | PHP |
| Tradebyte-CSV-To-XML-with-Python | Converts Tradebyte product exports from CSV to XML | Python |
| Zalando-Beauty-New-Products-Scraper | Scrapes and monitors new product arrivals | Python |
| Project | What it does | Stack |
|---|---|---|
| modernbert-ecommerce-categorization | Product categorization on e-commerce data using ModernBERT | Python, Jupyter |
- Agent-to-system integration: protocols (UCP / ACP), tool interfaces and authorization for AI agents
- Reliable data pipelines between marketplaces, shops and internal systems
- Applied ML on real-world business data, from experiment to production
Working on data integration, AI agents or applied ML? Reach out via ospeks.com or LinkedIn.
⭐ Pair Extraordinaire · Pull Shark · Quickdraw · Starstruck · Arctic Code Vault Contributor