Choose your path: a map of the deep-dive library

Time: about 15 min

---

The gist

Ahead of you is the deep-dive library, full of applied lessons. You won't read them all. This lesson is a map of where you should go. Answer 4 questions and you have your route.

Without this lesson, the applied lessons feel like a tool catalog: voice, video, analytics, CRM, support, ads. Everything is interesting, nothing gets built. A student reads the lessons one after another, remembers little two months later, and hasn't launched anything.

With this lesson, you leave with one of 5 ready-made routes: 8 to 12 lessons for your goal, in the right order, and a clear sense of why you're NOT reading the rest right now.

The course page also has three general paths by goal: the Use AI in my work path, the Earn with AI path and the Build my own product path. The map below is finer-grained: it picks a scenario inside the Earn with AI or Build my own product goal. If you took the Use AI in my work path, you've already finished the core course: you can simply skim this lesson.

🎨 Picture this: you've finished ski school. The green runs are behind you. In front of you is a mountain with dozens of trails. Which ones are yours? Without a trail map, you'll stand at the lift for an hour reading every name. This lesson is the trail map, color-coded: your level, your direction, your distance.

---

🎯 Decision tree: 4 questions to pick your route

The main question of this lesson is not "what's cool in the course" but "which of the applied lessons fit the way you plan to work with AI."

Answer the 4 questions in order. At the end you get one of 5 routes (A/B/C/D/E). The lessons that AREN'T on your route, you come back to later, when your scenario changes.

---

Q1: Who is your client?

If your answer is "both," pick the main scenario for the next 3 months. In 3 months, come back and take the second route.

🎨 Picture this: you can't drive to Chicago and to Dallas at the same time. One city first, then the other: months of work in each, not "both at once in a week."

---

Q2: What are you selling?

If you took the Earn with AI path, you've already covered some of these in the core course (the modules on niche, price and first clients). This is a reminder: don't skip connecting them to your new route.

---

Q3: Which stack (set of tools) are you betting on?

A stack isn't a route. The stack tells you which specific lessons to add to your route. The route is the sequence; the stack is the content.

---

Q4: How ready are you?

🎨 Picture this: you don't buy a Boeing 747 to run to the grocery store. And you don't buy a scooter to cross the Atlantic. Your readiness level means picking the right vehicle for the distance.

---

5 routes from Q1 to Q4

After the 4 questions, you're on one of 5 routes. Each route is a table: which lessons, in what order, how many weeks. You can skip any core-course lessons on your route that you've already taken.

---

Route A: B2B AI Agency (you sell AI services to businesses)

Client profile: companies with 10 to 200 people that are looking for automation and are ready to pay for a solution.

Route logic: clients → outreach → delivery → retention → packaging → scaling. Without a CRM and sales analytics, it doesn't work.

Time: 6 to 8 weeks, alongside real clients. Skip: Voice tools and Video editing (voice and video), the content factory lessons (AI copywriting–The content pipeline), and Product analytics–Competitive intelligence (analytics for your own product; you have a service, not a product).

---

Route B: B2C Consumer App (you sell an app to people)

Client profile: end users, a free version with paid extras (freemium) or a subscription; you need virality.

Route logic: MVP → UX → marketing → virality → retention → monetization. Mass scale means you need cheap infrastructure.

Time: 8 to 10 weeks. Skip: the B2B monetization lessons (Your first clients–The factory model), Voice AI agents–Chatbot managers (voice is premature for an MVP), and Niche and trend analysis and Competitive intelligence (competitive intelligence comes later).

---

Route C: Personal Automation (automating your own routine)

Client profile: you. Nothing is for sale; the payoff is indirect, in hours freed up for other work.

Route logic: find the routine → automate it → build up knowledge → scale it across all your projects. No public presence needed.

Time: 4 to 6 weeks (faster, since there's no outside client). Skip: the monetization lessons (Your first clients–The factory model; you don't need them), Cold outreach, Voice AI agents–Chatbot managers (voice is optional), and Product analytics–The final architecture (production is premature).

---

Route D: Corporate Consulting (you sell integrations to large companies)

Client profile: companies with 500+ people, an IT department, compliance requirements and pilot projects, ready to pay for reliability.

Route logic: the Microsoft ecosystem + security + audits + Teams. Without compliance and an audit log, you won't get past their review.

Time: 10 to 12 weeks (slower, because each lesson goes deep). Skip: Chatbots with the Claude API (a public chatbot isn't a corporate channel in most cases; large companies live in Teams), Voice tools and Video editing (voice and video can wait), the content factory lessons (AI copywriting–The content pipeline; not your core skill), and No-code AI–Zapier AI (no-code tools; corporations want custom work).

---

Route E: Content Creator (you earn through an audience)

Client profile: your audience (subscriptions, ads, sponsors, products). A YouTube channel, a newsletter or a blog as the base.

Route logic: content pipeline → virality → distribution across channels → monetizing the audience.

Time: 6 to 8 weeks. Skip: the B2B monetization lessons (Your first clients–The factory model), AI in messaging apps (that's messaging inside a company), Product analytics–Competitive intelligence (you'll need them, but later), and AI customer support–An AI chat on your website (customer support; you don't have clients in the traditional sense).

---

How to decide if you're torn between two routes

A real scenario: you went through Q1 to Q4 and ended up with two possible routes. For example:

• B2B agency (you know business owners who'd buy the service)
• Content creator (you've wanted a YouTube channel for a long time)

The rule for your first route: pick the one where money comes in faster.

These are rough guides for comparing routes, not an income forecast: results depend on your niche, your market and your work.

If your budget is tight → A or C. If you have 6+ months of runway → D or E.

🎨 Picture this: you're choosing between a vegetable garden (B2B: plant it, harvest in 2 months, money) and an apple orchard (content: plant it, it bears fruit in 3 years, but for longer and in bigger amounts). If there's food on the table today, plant the orchard. If not, dig the vegetable garden first.

---

Hybrid routes (after your first pass)

Three or four months after you finish your first route, you'll usually want to add a second stack. Typical combinations:

Never combine D and B at the same time: the rhythms are too different (6-month corporate cycles vs weekly B2C releases).

---

Not on any route (but good to know they exist)

These lessons are useful but not critical for most routes. Come back to them when your scenario calls for it.

---

❌ Common mistakes when choosing a route

❌ Reading all the applied lessons in a row over two months. After two months you remember little and haven't launched anything. That's catalog thinking: learning for the sake of learning.

❌ Jumping between routes at random. "Started route A, a week later jumped to E, two weeks after that to C." You lose the structure. A route is a sequence with a logic to it, not a pile of tags.

❌ Ignoring the production patterns lessons once you're in production. "My clients pay me, but I skipped plugin security and portfolio detachability." Sooner or later something breaks in production, and it's much harder to fix while a client is waiting. Once real clients depend on your work, we strongly recommend these lessons.

❌ Not picking a route at all ("I'm just reading"). Without a clear goal, the course turns into an encyclopedia. Encyclopedias are useful as a reference, not as a learning path.

❌ Picking 3 routes at once. "I'm doing a B2B agency, a B2C app and a YouTube channel." That's 24 weeks of focused work running in parallel. It won't happen. Pick ONE for 3 months and the second one after that.

❌ Ignoring Q4 (readiness level). Jumping into the production patterns lessons before your MVP is even live is overengineering. Skipping them once you're in production is underengineering. Your level has to match the distance.

🎨 Picture this: the course is a supermarket of tools. Without a shopping list, you wander the aisles for 3 hours, toss whatever looks shiny into the cart, and at home you can't figure out why you bought it. With a list, it's 20 minutes, and you walk out with what you're cooking tonight.

---

Coming back to the other lessons

A route is your first pass. Once you launch the first version (an MVP, a first client, a first paid product), you come back and take the remaining lessons, this time to solve specific problems.

Signs it's time to come back:

• "I realized I'm missing X" → read the lesson on X
• "A client asked for Y" → read the lesson on Y
• "My scenario changed (it was B2B, now it's B2C)" → take the second route all the way through
• "After a year of work, I want to scale" → go back to the production patterns lessons and to Advanced Claude Code tips for 2026

Bad reasons to come back:

• "Just curious": you'll spend hours and launch nothing
• "I have to take every lesson": that's catalog thinking again

---

Key takeaways

The applied lessons ahead aren't a queue to "read in order." They're a map of tools, and every tool has its own use. Pick the route that fits your scenario and come back to the rest when you need it.

5 routes cover the main ways to put Claude Code to work: a B2B agency, a B2C app, personal automation, corporate consulting, a content creator. If your scenario doesn't fit any of them, combine two, one after the other (3 months + 3 months).

The production patterns lessons are the line between "played around with Claude" and "built a production-grade system." Skipping them for an MVP is fine. Skipping them in production makes trouble much more likely once real clients depend on you.

This lesson isn't an instruction manual; it's a map. Come back to it every 2 to 3 weeks and check: am I still on my route, or am I drifting into "just curious"?

---

✅ Checklist

• Answered Q1: who my client is (B2B / B2C / B2Me / Corporate / Audience)
• Answered Q2: what I sell (a service / a product / content / data / SaaS)
• Answered Q3: which stack I'm betting on (text / voice / video / web / automation / data)
• Answered Q4: where I am (MVP / Beta / Production / Enterprise)
• Picked one of the 5 routes (A / B / C / D / E)
• Wrote down the 8 to 12 lessons on my route in a to-do app, Obsidian or a notebook
• Understand why I'm NOT reading the other lessons now (I'll come back to them when I need them)
• Know after which lesson on my route I go back to the production patterns lessons (if I'm in production)
• Set a start date for my route and an expected finish date (4 to 12 weeks, depending on the route)

---

Next lesson

The core course ends here. Next, follow your route from the tables above: open the first lesson on it that you haven't taken yet (you already know the core-course lessons on your route). Route A starts with Your first clients, B with Websites and web apps, C with An AI executive assistant, D with AI in messaging apps, and E with AI copywriting.

If you took the Use AI in my work path, you've finished the course: come back to the library when you have a specific task.