Skip to content

About this blog

Practitioner

Practitioner is a technical blog about AI-assisted engineering — the discipline of building production-grade software with the help of modern AI platforms and tools. The content spans the full landscape: Azure AI Foundry, GCP Vertex AI, Anthropic Claude, and the developer workflows that bring them together in real systems.

The articles here are not introductory tutorials. They are the guides you would write after shipping real projects — the setup decisions that save hours, the failure modes nobody documents, the architectural patterns that actually hold up under production load, and the practices that compound over time.

Topics include multi-agent architectures, LLM orchestration, mobile development with AI tooling, TypeScript workflows, cloud deployment, and the craft of building systems that other engineers can maintain and extend.

The goal is not to chase the latest model release. It is to understand what these platforms actually do well — and build accordingly.

Each article is written from direct production experience. If a technique is described here, it has been tested on real projects with real constraints — not just toy examples or conference demos.

What you will find here

  • Multi-agent system design

    Architecture patterns for building reliable agent pipelines on Azure AI Foundry, GCP Vertex AI, and with the Anthropic API — covering orchestration, state management, and quality gating.

  • AI-assisted development workflows

    How to integrate AI coding tools into real engineering workflows — from project configuration to automated quality pipelines — across different platforms and IDE setups.

  • Cloud platform deep-dives

    Practical guides for working with Azure AI services, Google Cloud AI, and Anthropic's APIs in production — including deployment, cost management, and safety configuration.

  • Production lessons

    The things you learn after shipping AI-powered features — the failure modes, the useful patterns, and the things worth doing differently next time.