AI for Makers and Solopreneurs

Descripción

You have hardware skills, a head full of ideas, and the ability to write a few lines of Python. What you need is a faster path from idea to working software. This course teaches hardware makers, electronics hobbyists, and solopreneurs to use AI coding assistants to build real software tools, end to end. You will learn not just how to prompt an AI, but how to run a complete development workflow: from brainstorming and specification through to a tested, deployed application. No computer science background required. Course structure The course runs across six modules. Module 0 covers installation and orientation. Module 1 builds the foundations: how AI-assisted coding actually works, the Five Levels framework, how to frame problems for an AI, and how to manage the failure modes that catch most new users. Module 2 covers the tools and LLM landscape. Module 3 walks through the full workflow: planning, implementation, testing, iteration, refactoring, and deployment. Module 4 covers embedding an LLM inside a running application. Module 5 is the capstone: a production-quality, multi-user web application built from scratch with an AI agent. Tools and LLMs The primary editor is VS Code. You will use two AI coding agents: Cline (a VS Code extension) and Claude Code (Anthropic's terminal-based agent). Both are agentic tools that read your files, write code, run commands, and iterate, with you directing the work. For language models, the course uses models from Anthropic (Claude), Alibaba (Qwen), and Moonshot AI (Kimi). You will see how to select and switch models based on the task and the cost. What you will build Four hands-on projects, increasing in scope and complexity: • Project 1 — RC Filter Analyser: An interactive browser tool with live Bode plots. Built using Claude in the browser only, no editor or terminal. Exercises the core framing skill from Module 1. • Project 2 — GUI Serial Monitor: A desktop application that reads live serial data from an Arduino and plots multiple variables in real time. • Project 3 — Datasheet Q&A Tool: Upload a component datasheet as a PDF, ask questions in plain English, and receive cited answers. Demonstrates the RAG pattern with an LLM running at application runtime. • Project 4 — Resource Booking System: A full multi-user web application for managing shared equipment in a lab or makerspace. Includes user authentication, booking conflict detection, public holiday integration via an external API, and email reminders. Optional extension: a natural-language booking agent. Solopreneur extension Module 5 includes an optional solopreneur track that extends the capstone into a shippable product. This covers multi-user data isolation, subscription payments with Paddle, transactional email with Resend, production deployment on a Hetzner VPS using Coolify, a product landing page, and user documentation. The maker track ends at a working, deployed application. The solopreneur track takes it all the way to a product people can pay for. Who this is for Hardware makers, PCB designers, Arduino and Raspberry Pi enthusiasts, and solopreneurs who want to build and ship real software. You should be comfortable writing a few lines of Python. No application architecture or AI tooling experience is needed.

Lo que aprenderás

• Use AI coding agents (Cline and Claude Code) to take a project from first idea to a working, deployed application.
• Write prompts that give an AI enough context to produce useful, correct output on the first attempt.
• Explain the Five Levels of AI-assisted coding and choose the right level and tool for any given task.
• Recognise and manage the six most common failure modes of AI-generated code before they derail a project.
• Build four hands-on projects: a browser tool, a GUI desktop app, an LLM-powered Q&A tool, and a full web application.
• Run a complete AI-assisted workflow: brainstorm, plan, implement, test, iterate, refactor, and deploy.
• Embed a large language model inside a running application to answer questions from a PDF datasheet.
• Extend a working application into a shippable product with authentication, payments, and deployment.

Requisitos

• Comfortable writing simple Python scripts — no advanced Python or software engineering experience required.
• Familiar with at least one microcontroller platform, such as Arduino or Raspberry Pi.
• A computer running Windows, macOS, or Linux with an internet connection.
• VS Code installed (free; full installation walkthrough provided in Module 0).
• An API key for at least one LLM provider: Anthropic, Alibaba (Qwen), or Moonshot AI (Kimi). Free tiers are available for all three.
• No prior experience with AI coding tools, application architecture, or deployment is needed.