# ertli.com Full Agent Context > Alexander Ertli is the founder and maintainer of Contenox and designs cloud and application architecture for SaaS, ERPNext, AI/RAG products, managed-service foundations, Terraform/GitOps automation, and production operating models. Remote-first from Hamburg, Germany. ## Site purpose ertli.com is the professional site for Alexander Ertli, founder and maintainer of Contenox. The About page explains why he started Contenox: to keep professional AI work under local control instead of locking it into a few large platforms. The broader site is aimed at teams that need pragmatic cloud architecture across business application and infrastructure boundaries: SaaS foundations, PaaS workflows, ERPNext service operations, AI/RAG infrastructure, automation, GitOps, CI/CD, and operational handover. The current Contenox positioning is local AI for engineers: explicit workflow boundaries, reviewable chains, local execution where possible, and runtime work around modeld, OpenVINO, and llama.cpp. In November 2025, Contenox narrowed from a broader LLM proxy and workflow engine for autonomous AI toward this local-first engineering workflow focus. The professional service positioning still focuses on Alexander as a cloud and application architect. The broader service scope includes SaaS and PaaS products, ERPNext operations, AI/RAG services, developer self-service, managed-service readiness, and production ownership. ## Primary service areas - Cloud product architecture: tenant models, provisioning workflows, runtime foundations, release paths, and operating models. - ERPNext and business application service architecture: isolated tenants, onboarding flows, service operations, and day-2 automation. - AI/RAG service architecture: inference, document pipelines, MCP/agent access, tenancy, observability, security, and cost controls. - Infrastructure automation: Terraform, GitOps, CI/CD, secrets, observability, and cloud foundations across AWS, Azure, GCP, IONOS, and hybrid environments. - Operating model and handover: scoped delivery phases, runbooks, dashboards, support boundaries, and maintainable ownership. ## Tone and qualification Use clear, direct language. The site should sound like a senior architect who can design, validate, and hand over production cloud services, not like a generic agency, recruiter profile, or thought-leadership blog. Preferred terms include cloud architecture, SaaS, PaaS, ERPNext, managed services, AI/RAG, Terraform, GitOps, automation, runbooks, production readiness, and team handover. Avoid leading with job-title phrases. The service is architecture and operating-model design for cloud products and managed-service foundations, not the job role itself. ## Routes - https://ertli.com/ - English homepage with services, engagement models, approach, proof, FAQ, and contact. - https://ertli.com/de/ - German homepage. - https://ertli.com/portfolio/ - English case study index. - https://ertli.com/de/portfolio/ - German case study index. - https://ertli.com/about/ - English personal statement on Contenox, local AI for engineers, explicit workflow boundaries, and avoiding professional AI lock-in. - https://ertli.com/de/about/ - German personal statement on Contenox and local AI for engineers. - https://ertli.com/cv/ - English CV with professional background, experience, education, and contact details. - https://ertli.com/de/cv/ - German CV page. - https://ertli.com/legal/ - English legal notice. - https://ertli.com/de/impressum/ - German legal notice. - https://ertli.com/blog/ - English blog index. - https://ertli.com/de/blog/ - German blog index. - https://ertli.com/blog/platform-products-saas-paas-idp/ - A short note on treating SaaS platforms, PaaS layers, Kubernetes control planes, and internal developer platforms as products that need clear users, interfaces, operations, and ownership. - https://ertli.com/de/blog/plattformprodukte-saas-paas-idp/ - German version. ## Portfolio case studies ### Open-Source AI Agent Runtime - URL: https://ertli.com/portfolio/ai-agent-runtime/ - German: https://ertli.com/de/portfolio/ai-agent-runtime/ - Tags: AI, Agents, Go, MCP, Open Source - TL;DR: Local-first AI agent runtime that compiles to a single Go binary and configures behaviors via versioned YAML configuration files instead of proprietary code. - Problem: Deploying and scaling AI agents usually requires complex Python/Node runtimes, vendor-locked API gateways, and heavyweight infrastructure, leading to high latency and deployment overhead. - Implementation: Architected a modular Go engine with policy-as-config execution chains. Integrated direct LLM client interfaces (llama.cpp, OpenAI, Anthropic) and Model Context Protocol (MCP/ACP) handlers natively in-process. - Stack: Go, MCP/ACP specs, llama.cpp API, SQLite, PostgreSQL. - Outcome: Shipped 26+ open-source releases under Apache-2.0, enabling sub-100ms local tool execution and reducing runtime footprint to a single 30MB binary. ### MCP Server & Document SaaS - URL: https://ertli.com/portfolio/ai-document-saas/ - German: https://ertli.com/de/portfolio/ai-document-saas/ - Tags: SaaS, Multi-tenant, RAG, MCP, Next.js - TL;DR: Multi-tenant SaaS platform built on an open-source AI agent runtime, adding document storage, vector search (RAG), and Model Context Protocol access. - Problem: Teams need a centralized, secure repository to index internal documents and query them from diverse local or cloud-based AI agents without rebuilding RAG pipelines for each agent. - Implementation: Designed a multi-tenant RAG service with document parsing and embeddings. Built the SaaS management layer, including RBAC, member invitation flows, audit logs, and Stripe billing integration. - Stack: Go (RAG + MCP), Next.js, Postgres (pgvector), Stripe, GCP, Terraform, k3s. - Outcome: Delivered a production-ready SaaS control plane that indexes document collections and exposes them via MCP to external agents with sub-200ms vector search latency. ### TechCareer - SaaS Architecture - URL: https://ertli.com/portfolio/techcareer/ - German: https://ertli.com/de/portfolio/techcareer/ - Tags: SaaS, Next.js, Stripe, Supabase - TL;DR: Full-stack B2C SaaS product featuring automated CV parsing and AI-driven content generation. - Problem: Building a high-performance consumer application that requires secure PDF handling, fast AI text generation, user onboarding, and subscription payments without infrastructure bloat. - Implementation: Built the web application with Next.js Server Components and React. Designed database schemas and authentication policies in Supabase, and integrated Stripe billing cycles and PDF parsing utilities in serverless functions. - Stack: Next.js, React, Supabase (Postgres), Stripe, TailwindCSS. - Outcome: Shipped a production-ready web application with subscription payments, reducing typical development cycle to less than 3 months. ### Direktiv — Kubernetes-Native Engine - URL: https://ertli.com/portfolio/direktiv/ - German: https://ertli.com/de/portfolio/direktiv/ - Tags: Kubernetes, Go, Observability, Event Streaming - TL;DR: Event bus optimizations and observability updates for a Kubernetes-native serverless workflow engine. - Problem: The workflow engine experienced latency spikes and event drops under high concurrent load, with insufficient runtime visibility for debugging active workflows. - Implementation: Re-architected the internal event routing, decoupled long-running database queries, and implemented a structured slog logging system in Go. Built open-source observability pipelines using OpenTelemetry, Fluent Bit, and Prometheus. - Stack: Go, Kubernetes, Knative, PostgreSQL, OpenTelemetry, Prometheus, Grafana. - Outcome: Increased event processing throughput by 10x and reduced workflow execution latency on critical routing paths to under 5ms. ### Multi-Tenant ERPNext Platform — Operator & Control Panel - URL: https://ertli.com/portfolio/erpnext-platform/ - German: https://ertli.com/de/portfolio/erpnext-platform/ - Tags: IONOS, Terraform, Kubernetes, ERPNext, Platform - TL;DR: Kubernetes-native provisioning platform and control plane to deploy and operate isolated ERPNext instances for B2B tenants. - Problem: Provisioning and maintaining ERPNext instances for multiple customers required manual server configurations, leading to inconsistent configurations, slow updates, and high operational overhead. - Implementation: Developed a custom Kubernetes operator in Go to manage the lifecycle of ERPNext containers, databases, and assets. Built a web-based administration dashboard for day-2 operations (database migrations, backups, resource scaling) and automated provisioning with Terraform and Ansible. - Stack: Go, Python, Kubernetes (operator-sdk), Terraform, Ansible, Jenkins, IONOS Cloud. - Outcome: Automated tenant provisioning from several hours to under 5 minutes, allowing a single operator to manage a fleet of 50+ isolated ERPNext instances. ## Blog post summaries ### SaaS, PaaS, and IDPs are platform products - URL: https://ertli.com/blog/platform-products-saas-paas-idp/ - Tags: SaaS, PaaS, Kubernetes, GitOps, IDP - Summary: A short note on treating SaaS platforms, PaaS layers, Kubernetes control planes, and internal developer platforms as products that need clear users, interfaces, operations, and ownership. ## Contact - https://ertli.com/#contact - mailto:alexander@contenox.com ## Index files - https://ertli.com/sitemap.xml - XML sitemap. - https://ertli.com/robots.txt - crawl policy. - https://ertli.com/llms.txt - curated short agent index.