DRIP.

The roadmap  ·  Zain Qurashi

Nine levels from zero to AI integrations engineer.

Why integrations: with nothing connected, AI can only talk. Connected to a company's systems, safely, it does real work, and connecting it is the job. Every level is a real thing you build, not a quiz. Levels 0, 1, 2 and 4 are free. The Lab adds the full builds, a plan for your level and live help. Print it, or keep it open while you work.

  1. 0

    Start here: build your own dictation tool Free

    You learnWhat an API key is and where it lives. How to describe a tool to an AI and get a working one back. How to measure what you built.

    You buildA dictation tool, tonight, in your browser. Then the same tool again, better, as you go. Open the lab

  2. 1

    Curious: how AI talks to software Free

    You learnWhat an API is. A prompt is not a question. Keeping keys safe. Tokens, cost, and model vs effort.

    You buildA script that calls a model, with every line explained.

  3. 2

    Connected: AI that decides Free

    You learnDecision models: yes or no, pick one, score it. Why context changes the answer.

    You buildA tool finder that ranks 200 AI tools for a typed task. It runs this site's directory.

  4. 3

    Builder: connect AI to your apps Lab

    You learnAPIs, MCP and connectors. What an AI can and can't touch once it's plugged in.

    You buildYour first MCP server, so an AI can use an app of your choosing.

  5. 4

    Integrator: voice and media Free

    You learnHow voice to text works, end to end, and where it breaks.

    You buildVersion 2 of your dictation tool, on your own computer. The full Scribe app is the Lab build.

  6. 5

    Automator: automations and agents Lab

    You learnHow a company rolls out Claude and ChatGPT: which plan, SSO, which connectors may read and which may write, the AI policy. The automations every company runs. When an agent is worth it, from no-code to code.

    You buildA company on real tools (Okta, Slack, Airtable, HubSpot, QuickBooks), then ten automations in it: new hires and leavers by group, access requests approved in Slack, email into Airtable, bill approvals, the Monday report. Then its agents.

  7. 6

    Shipper: AI at the enterprise level Lab

    You learnWhy a chat license is not an AI strategy. Agent identity, read versus write permissions, human approval for writes, logging every agent action, MCP proxies, prompt injection, cost limits, and what a security team asks before saying yes.

    You buildIn the same company: a read-only MCP server, approvals for writes, an agent that acts as the user, your own MCP proxy, logs of everything, attacks you fix, a Copilot and an AWS agent behind the proxy, and the rollout review.

  8. 7

    Engineer: portfolio projects Lab

    You learnHow to scope, ship and write up a project someone else can run.

    You buildFour projects to show an employer or a client.

  9. 8

    Architect: automating a character Case study

    You learnThe Drip pipeline: script, voice, animation, posting. And the legal side: AI disclosure, who owns an AI character, voice and likeness rights, sponsorship rules.

    You buildThe same pipeline, the way a company would run it.

  10. 9

    Legend: the career Lab

    You learnWhat the job is day to day. Portfolio, resume, interviews. Your first freelance automation client.

    You buildA plan for your next twelve months, reviewed live.

  11. +

    Side quest: record, edit and make videos with AI Lab

    You learnScripts, AI voices, characters animated in code, and the honest part: disclosure and voice rights.

    You buildYour own recording and edit setup. Members get Take, my screen recorder.