Skip to main content

Stack

A stack organized around outcomes.

Tools matter when they reduce delivery risk, connect cleanly, and remain understandable after launch.

Editorial study of product technology layers
01

Reasoning and orchestration

AI & Automation

  • OpenAI
  • Gemini
  • LangGraph
  • RAG
  • n8n
  • Python

Used for agent workflows, retrieval, classification, extraction, and connected business processes.

02

Interfaces and services

Web Runtime

  • Astro
  • Next.js
  • React
  • Laravel
  • REST APIs

Chosen around content, interaction, and deployment needs rather than framework novelty.

03

Cross-platform delivery

Mobile

  • Flutter
  • Dart
  • Firebase
  • Dart FFI

For mobile products that need one product surface with access to native performance when required.

04

Structured operating knowledge

Data & CMS

  • PostgreSQL
  • Redis
  • Directus
  • Strapi
  • WordPress

Keeps business data editable, queryable, and available to both product and automation layers.

05

Production path

Infrastructure

  • Docker
  • Linux VPS
  • Nginx
  • GitHub

A small operational stack for repeatable deployment, observability, and maintainable handoff.

06

What is being tested

Current experiments

  • Agent evaluation
  • Document intelligence
  • Local inference

Exploring reliable evaluation loops and lower-cost ways to run useful AI features.

Use the smallest stack that can carry the system.

Every additional service adds another failure mode. I add layers when they earn their place through capability, reliability, or speed of delivery.

Your next project

Have a project in mind?

Tell me what you want to automate, build, or improve.