Skip to content
View manuelblinkert's full-sized avatar
🎯
Focusing
🎯
Focusing

Block or report manuelblinkert

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
manuelblinkert/README.md

Manuel Blinkert

AI Systems Architect · Founder/CTO · Mathematician

I build the system around the model: architecture, backend, data, evaluation, observability, and cost control for AI products that have to hold up in production.

M.Sc. Mathematics (TU Berlin), 8+ years in software engineering and data science, CTO and co-founder of a stealth AI startup.

What I Work On

  • Architecture for AI products that outgrow the prototype
  • LLM orchestration, agent workflows, and context design
  • Python and FastAPI backends, Postgres, async workers
  • Evaluation, observability, reliability, and cost control
  • Memory, state, and knowledge infrastructure

Selected Public Work

A FastAPI starting point for LLM services you can run and debug in production, with route/service/schema separation, provider abstraction, structured JSON logging, request IDs, tests, Docker, and CI.

Use it to find the weak points of an AI system before launch. It covers architecture, LLM workflows, observability, evals, RAG and context design, cost control, security, deployment, and reliability.

An authenticated remote MCP server that lets Claude, ChatGPT, and other AI apps read and write a single GitHub-backed knowledge vault. Scoped tool registry, token-based auth, per-member commit attribution, and a hard single-repository boundary.

The design and reasoning: Beyond Chat History: Building an AI-Native Second Brain with Obsidian, GitHub, and MCP.

Architecture notes and a review template on workflow orchestration, semantic observability, and system boundaries in production AI systems.

Writing and Teaching

I write about how modern software and AI systems work and how to make good engineering decisions about them: mechanisms, trade-offs, and what breaks in practice. Articles are on my blog.

I also teach AI software engineering to people who want to turn an idea into a working product.

Work With Me

I take on selected freelance projects in AI and software architecture, focused on the DACH region and the rest of Europe. Most start with a fixed-scope reliability review of an existing AI product.

Book a 20-minute fit call or read more on my website.

Tech Stack

Python · FastAPI · TypeScript · React · Supabase · Postgres · LangGraph · LangChain · OpenAI SDK · Docker · DigitalOcean · Grafana · PySpark · Databricks

Contact

Pinned Loading

  1. ai-backend-production-starter ai-backend-production-starter Public

    A production-minded FastAPI backend starter for AI/LLM applications with structured logging, request IDs, clean service boundaries, tests, and Docker.

    Python 1

  2. ai-production-readiness-checklist ai-production-readiness-checklist Public

    A practical checklist for evaluating production readiness of AI systems: LLM workflows, architecture, observability, evals, cost, reliability, RAG, deployment, and security.

  3. second-brain-github-mcp second-brain-github-mcp Public

    Authenticated remote MCP server for Claude, ChatGPT, and other AI apps to safely read and write one GitHub-backed second-brain vault — single-user or team, with per-member commit attribution.

    Python 4 1

  4. ai-systems-architecture-notes ai-systems-architecture-notes Public

    Practical architecture notes on production-grade AI systems, LLM orchestration, observability, backend architecture, and reliability.