2-Week AI Builder Intensive

MakerSquare's 2-Week AI Builder Intensive teaches non-technical professionals — operators, founders, lawyers, marketers, finance and ops professionals — to build and deploy working AI tools using production software, not no-code platforms. It runs in person in Austin, TX, Monday through Friday from 9am to 3pm for two weeks in small cohorts, and every student builds on their own real data and workflows starting Day 1. Students build a live website or internal dashboard backed by a real database, automations running on their actual work tools, and a working AI agent deployed on a schedule — learning along the way when a problem calls for an agent versus a simple automation — then present a capstone on stage at a public Demo Day at Capital Factory. The stack is production software used in the real world: Claude and Claude Code, Warp, GitHub and GitHub Actions, Lovable / Figma Make, Vercel, Supabase, Composio, Exa, Render, and Obsidian. No coding background is required.

• Before You Learn / Who This Course Is For

  • Operators and business owners. You know which half of your week should be running itself. Knowing what to automate was never the hard part for you — building it is what wasn't available. Now it is.

  • Professionals and career builders. Marketing, finance, healthcare, law, recruiting. Domain expertise plus the ability to build with AI is the rarest profile in your industry, and it's a moat that doesn't close.

  • Founders and first-time builders. You don't need a technical co-founder to ship the first version. If you can describe the product clearly, you can build it.

  • Doctors, lawyers, marketers, recruiters, and operators have all built here. No coding, no math, no computer science. What you need is basic computer fluency and a real problem from your own work — that problem is what you build on starting Day 1.

  • No interview, no essay, no technical test. Fill out the form, pay tuition, your seat is confirmed within 24 hours. We trust you to know whether this is right for you.

  • Not the right fit if…

    • You're changing careers into a software engineering job. This is not a job-placement program and we don't train developers.
    • You want a certificate more than you want working tools.

    • You want self-paced videos instead of a room full of builders.

    • You can't commit two full weeks in Austin.

    • You'd rather study the theory than ship something.

• What to Expect

  • Two weeks, in person, on your own work. Monday to Friday, 9am to 3pm, at Capital Factory in Austin. Small cohorts. Afternoons stay yours — the program is built so an urgent thing at work doesn't cost you the day.

  • You bring a real problem and real data from your own work, and you build on it from Day 1. Nothing here runs on a toy dataset or a shared example project. What you build is yours, and it's still running when you get home.

  • Week 1 — Build. How models actually work, and how to prompt like an expert instead of a search engine. Your builder stack: the terminal, GitHub, Claude Code, and a CLAUDE.md that makes Claude work the way you work. Then you build the web thing your project actually needs — a public site or an internal dashboard — and get it live. Then a real database behind it: a live form writing real rows, and the safety rules that keep your data yours.

  • Week 2 — Automate and ship. Agents start Monday. Everyone builds one that runs on a schedule and keeps arriving in your inbox long after the program ends. Then you build your own — automations wired into your real Gmail, Slack, and sheets, or a coded agent, whichever your problem actually calls for. Thursday you make it hold up under real use. Friday is Demo Day.

  • Demo Day. Five minutes on stage at Capital Factory, live demo on real data, in front of Austin builders, founders, operators, and investors. Real questions, real feedback.

  • The tools — every one used in production:

    • Claude — your primary model

    • Claude Code — a full project collaborator in your terminal

    • Warp — the terminal, made usable

    • GitHub + GitHub Actions — version control, and the scheduling that keeps your agent running

    • Lovable / Figma Make — how you build the site or dashboard;

    • Vercel — one-click deployment

    • Supabase — database and backend

    • Composio — connects Claude to Gmail, Slack, and 200+ tools

    • Exa — search, as your agent's eyes

    • Render — hosting for agents and automations

    • Obsidian — your builder's knowledge base

• What You'll Achieve

  • Deployed, working tools. A live website or internal dashboard at a public URL, backed by a real database. Automations running on your actual work tools. An AI agent on a schedule that keeps doing its job after you leave. All built on your real data — yours to keep using Monday morning.

  • Judgment, not just tools. When a problem needs an agent and when it just needs a well-built automation. Most real problems don't need an agent. Knowing which ones do is what separates people who ship from people who over-engineer, and almost nobody teaches it.

  • Fluency in the real stack. The terminal, version control, deployment, databases, API keys and where they live. Not so you become an engineer — so that no part of building is a black box you have to hand to someone else.

  • How to work with Claude as a collaborator. Expert prompting, CLAUDE.md, driving the conversation yourself instead of pasting someone else's prompts. This is the skill that compounds every month after you leave.

  • Proof. A Demo Day presentation in front of Austin's tech community. A live URL you can show a client, a boss, or an investor. And a certificate of completion for LinkedIn.

  • A cohort. A small group of builders who stay your network, plus lifetime access to the alumni Slack and every future Demo Day.