Production systems · 2024–present
AI Agents in Production
Autonomous AI agents with LLM integration in production web applications — from automating internal workflows to real-time user interaction.
- LLM
- LangChain
- Go
Fullstack Software Engineer — Munich
I build scalable microservices and AI agents at Check24 Factory. Here I show both sides: how I architect systems — and why they stay invisible to both humans and AI models without SEO and llms.txt.
Fullstack Software Engineer focused on scalable microservices and high-throughput web applications — built with Go, Node.js and PHP (Symfony), complemented by modern frontend frameworks. Currently at Check24 Factory, responsible for architectural decisions, asynchronous message queuing, and AI-driven workflow automation.
Previously: end-to-end development at Vivid Planet Software and frontend work at Eyedea Werbe. B.Sc. in MultiMedia Technology (FH Salzburg), focused on web technologies and human-computer interaction — where my interest in how systems get read by humans and machines started.
The toolbox behind the projects below.
Case studies from production systems — anonymized, but concrete.
Production systems · 2024–present
Autonomous AI agents with LLM integration in production web applications — from automating internal workflows to real-time user interaction.
High-traffic system
Decoupled, fault-tolerant architectures using RabbitMQ & Kafka for reliable asynchronous processing under high load.
Enterprise platform
Integrating Elasticsearch & Solr into existing systems — from relevance tuning to scaling across large datasets.
Developer experience
Designed and built a specialized test system inside the Bitbucket pipeline to speed up internal engineering workflows.
My second specialty
The best code is worthless if nobody finds it — neither humans via Google, nor assistants like ChatGPT, Claude or Perplexity. That's my second field: building systems that stay readable, citable and discoverable for both search engines and language models.
Crawlability, Core Web Vitals, semantic HTML and structured data (Schema.org / JSON-LD) still decide whether a page shows up in Google at all.
llms.txt is an emerging standard — akin to robots.txt and sitemap.xml — that gives AI systems a curated, plain-markdown overview of a site's key content. No navigation, ads, or JS rendering to slow crawlers down.
This page practices what it preaches: semantic HTML, a Person schema (JSON-LD) in the source, a bilingual hreflang setup — and a real, live /llms.txt.
# Matthias Riedl> Fullstack Software Engineer (Go, Node.js, PHP/Symfony, React) focused on> microservices, AI agents and LLM integration at Check24 Factory, Munich.## Site- [Home](https://matthiasriedl.de/): overview, skills, projects, contact- [Home (English)](https://matthiasriedl.de/en/): English version of this site## Projects- [AI Agents in Production](https://matthiasriedl.de/#projects): autonomous LLMagents in production web applications- [Event-Driven Microservices](https://matthiasriedl.de/#projects): Kafka/RabbitMQarchitectures for high-traffic systems- [Search Infrastructure](https://matthiasriedl.de/#projects): Elasticsearch/Solrintegration for enterprise platforms## Legal- [Legal Notice](https://matthiasriedl.de/impressum/)- [Privacy Policy](https://matthiasriedl.de/datenschutz/)## Contact- Email: info@matthiasriedl.de- LinkedIn: https://www.linkedin.com/in/matthias-riedl-68a18819b/
Contact
Project inquiry, technical question, or a heads-up about a broken llms.txt — write to me.
info@matthiasriedl.deLinkedIn